Abstract
In the current economic landscape, particularly in China, the development of new quality productive forces (NQPF) has emerged as a strategic approach for driving innovation and advancing sustainable productivity. Using a panel dataset of Chinese A-share listed firms spanning 2011 to 2023, this study empirically examines the influence of digital technology (DT) on the evolution of NQPF and explores its underlying transmission mechanisms and heterogeneity among different enterprises. The findings demonstrate that DT significantly contributes to the advancement of NQPF, a result that remains consistent under extensive robustness and endogeneity checks. Mechanism testing reveals that green technological innovation and enhanced managerial efficiency serve as critical mediating channels through which DT exerts its impact. Furthermore, the moderating role of market competition intensity is found to meaningfully shape this relationship. The heterogeneity analysis indicates that DT’s positive influence on NQPF is particularly strong among state-owned, low-pollution, and high-tech enterprises, as well as firms located in eastern regions and operating in technology-intensive industries. This research enriches the theoretical discourse on the digital economy and productivity transformation by providing enterprise-level empirical evidence to guide digital strategy formulation and policy design.
Plain Language Summary
This study examines how digital technologies help businesses improve their productivity in more sustainable and efficient ways. We analyzed data from Chinese publicly-traded companies over a 13-year period and found that companies using more digital technologies tend to perform better in terms of environmental sustainability, operational efficiency, and innovation capacity. Specifically, digital technologies help companies develop greener innovations and manage their operations more effectively. These positive effects are particularly strong for companies facing intense market competition, located in eastern China, operating in high-tech industries, or being state-owned enterprises. Our findings suggest that policymakers and business leaders should consider supporting digital transformation efforts, especially for small and medium-sized enterprises and companies in less developed regions, while being mindful of the different needs and capabilities across various types of businesses.
Keywords
Introduction
In recent years, the pursuit of sustainable and innovation-driven economic growth has brought increasing attention to the drivers of productivity that balance economic advancement with ecological sustainability. In China, this priority is embodied in the national strategy of developing new quality productive forces (NQPF). First proposed by President Xi Jinping in September 2023, NQPF emphasizes actively cultivating strategic emerging and future industries to foster new engines of high-quality economic growth (Xue & Chen, 2025). Although this concept is rooted in China’s contemporary policy context, it aligns closely with broader international interests in green total factor productivity and low-carbon innovation, as it aims to promote sustainable development through resource-efficient and environmentally friendly technological progress (George et al., 2020).
At the policy level, the importance of NQPF has been repeatedly underscored. In March 2025, the National People’s Congress (NPC) and the Chinese People’s Political Consultative Conference (CPPCC) reiterated the strategic role of technological innovation in industrial transformation, particularly through the adoption of disruptive and frontier technologies (Yang & Che, 2025). This focus aims to cultivate emerging industries, foster innovative business models, and stimulate new growth drivers, thereby accelerating NQPF’s rapid development (Ren et al., 2025). Guided by innovation, NQPF aims to achieve sustainable development by minimizing environmental impact and enhancing resource efficiency through technological progress, thereby fostering a harmonious balance between economic growth and ecological sustainability—aligning closely with the principles of green productivity (J. Zhang & Liu, 2025).
Meanwhile, the global economy is experiencing a paradigm shift driven by digitalization. Digital transformation (DT)—enabled by technologies such as the Internet, big data, cloud computing, and the Internet of Things—has become a key engine of productivity improvement and industrial upgrading (Cirillo et al., 2024; Sayar et al., 2025). These technologies reshape production methods, optimize resource allocation, and enhance innovation performance, contributing to the emergence of new forms of productivity including network, information, and data productivity (Ding et al., 2025). Empirical evidence further suggests that DT significantly facilitates enterprise innovation, operational restructuring, and sustainable value creation (Lokuge et al., 2025; Stein & Prost, 2024). Therefore, DT is expected to function as a fundamental enabler for enterprises to enhance their productive capabilities and promote the cultivation of NQPF.
The comprehensive and sustained integration of DT with the real economy has acted as a powerful catalyst for the rapid emergence and robust development of NQPF, accelerating the digital transformation and modernization of enterprises. The deep integration of digital technologies with modern industrial systems not only fosters a more efficient and sustainable digital economic ecosystem but also provides a solid foundation for advancing NQPF (H. Liu & Li, 2025). By leveraging digital tools, enterprises can promote green transformation, achieve carbon reduction throughout entire processes, and enhance the efficient allocation of social resources. Moreover, DT facilitates not only the intelligent upgrading of internal production and manufacturing systems but also the optimization of management processes, thereby improving overall management efficiency.
This raises several critical questions: How does DT significantly influence the enhancement of NQPF within enterprises? What evolutionary mechanisms underlie this impact? Do these effects vary significantly across different types of enterprises? Addressing these key questions is essential for developing a comprehensive understanding of the mechanisms by which DT drives the formation and growth of NQPF, and for providing both theoretical insights and practical evidence to support its effective enhancement.
To empirically examine the effect of digital transformation on the development of new quality productive forces, this study constructs a panel dataset of Chinese A-share listed manufacturing firms from 2011 to 2023. Digital transformation is measured through a text-mining approach based on annual reports, while NQPF is evaluated using a multi-dimensional index system capturing innovation capability, sustainable efficiency, and digital productivity. A series of econometric models are developed to estimate the causal relationship between DT and NQPF, complemented by mechanism analyses and heterogeneity tests. The empirical results reveal that DT significantly promotes the formation and enhancement of NQPF, and the effect remains stable after multiple robustness checks, including alternative variable measures, lagged estimation, and placebo tests. Furthermore, green technology innovation and management efficiency are identified as key mediators, while market competition plays a significant moderating role. These findings provide systematic evidence clarifying how DT contributes to NQPF development in China’s manufacturing sector.
This study makes contributions by addressing the research questions from three perspectives. By undertaking an investigation at the enterprise level, it first examines how Digital Transformation (DT) influences the development of NQPF, thereby enriching the discourse on DT’s economic consequences and offering insights for adjacent research fields. Second, in contrast to the predominantly theoretical and empirically limited nature of existing literature, this study employs a panel dataset of A-share listed companies from 2013 to 2023 to furnish robust empirical evidence. Third, it delves into the underlying mechanisms, aiming to unpack the “black box” of the DT-NQPF relationship. The analysis specifically identifies green technology innovation and management efficiency as internal mediators, while also considering market competition as an external moderator. This multifaceted approach elucidates the transmission pathways, enhancing the theoretical framework for understanding how DT dynamically shapes NQPF.
Literature Review
A growing stream of literature has begun to explore the nexus between DT and NQPF. This emerging body of work can be broadly synthesized into two key themes: (a) the conceptualization and measurement of NQPF, and (b) the theoretical and empirical examination of the relationship between digitalization and NQPF.
First, in terms of conceptual understanding and measurement, scholars have examined the driving forces and structural composition of NQPF. It is widely acknowledged that NQPF is shaped by the interplay of technological breakthroughs, factor innovation, and industrial upgrading, reflecting the integration of technological innovation and industrial transformation, which collectively optimize production methods and foster high-quality economic development (J. Li et al., 2024). This conceptualization resonates with the broader international discourse on productivity, which highlights the role of general-purpose technologies such as AI and robotics in driving productivity growth (Brynjolfsson & Hitt, 2003). However, while the latter focuses on automation and labor substitution, NQPF places greater emphasis on the synergistic evolution of laborers, means of labor, and objects of labor, integrating green and digital dimensions. Additionally, the multidimensional nature of NQPF—encompassing technological, industrial, and factor-related aspects—has been analyzed to highlight its profound influence on economic progress (Shao et al., 2024). Additionally, the multidimensional nature of NQPF—encompassing technological, industrial, and factor-related aspects—has been analyzed to highlight its profound influence on economic progress (Feng et al., 2024). This aligns with the OECD’s emphasis on multi-dimensional productivity frameworks that incorporate environmental and social outcomes, though NQPF extends this by embedding a stronger focus on systemic transformation of the production system itself.
Parallel to conceptual discussions, the literature has seen the emergence of diverse quantitative approaches for evaluating NQPF. One stream of scholarship builds comprehensive indicator systems incorporating technological, green, and digital productivity, with methods like the improved entropy weight-TOPSIS model being applied to measure provincial NQPF in China (L. Li & Liu, 2024). Another approach adopts an index system incorporating new quality laborers, means of labor, and objects of labor, using models such as projection pursuit to evaluate NQPF at the prefecture-level city scale (Huang et al., 2024). Together, these studies furnish essential methodological foundations for empirically assessing the progression of NQPF.
Second, regarding the relationship between digitalization and NQPF, most existing literature focuses on theoretical discussions. It is generally agreed that digital transformation plays a crucial role in promoting the deep integration of the digital and real economies, thereby accelerating the cultivation and strengthening of NQPF and supporting high-quality economic development (Yao et al., 2025). Some studies further examine the enabling role of digital infrastructure, management experience, and production models, suggesting that these elements offer essential technical support, innovative management concepts, and effective reference models for enhancing NQPF (S. Chen & Alexiou, 2025). In addition, the integration of DT is considered a key driver for breakthrough innovations, facilitating transformative changes in productivity and industrial structure (H. Liu & Li, 2025). From an empirical perspective, however, relatively few studies investigate the micro-level effects of digitalization on enterprise-level NQPF. The limited available evidence suggests that digital transformation can improve NQPF by strengthening human capital, alleviating financial constraints, and enhancing innovation capacity, ultimately optimizing resource allocation and improving competitiveness (J. Zhang & Liu, 2025). Other empirical findings demonstrate that digital transformation promotes NQPF through technological and managerial innovations, which elevate enterprise technological capabilities and refine management models, leading to comprehensive productivity gains (Y. Xu et al., 2025).
In summary, while the academic literature has made notable progress in advancing the theoretical understanding of NQPF and its macro-level implications, research remains relatively scarce in terms of empirical analysis and measurement at the enterprise level. Addressing this gap is essential for deepening our understanding of how digital transformation can effectively enhance NQPF within firms.
Theoretical Analysis and Hypotheses
DT and NQPF of Enterprise
For the purpose of this study, we conceptualize NQPF as an advanced configuration of productivity, marked by its technological intensity, superior efficiency, exceptional quality, and inherent sustainability. It represents a qualitative leap driven by the synergistic evolution and optimized integration of the three core elements of production: laborers (e.g., a digitally skilled workforce), means of labor (e.g., intelligent machinery and AI), and objects of labor (e.g., new energy and advanced materials). While NQPF encompasses the goals of green total factor productivity and is powerfully enabled by digital transformation, it is distinct in its holistic focus on the transformative upgrade of the entire production system itself, rather than any single performance outcome. Schumpeter’s theory of innovation and endogenous growth theory provide the theoretical underpinnings for this analysis, positing that technological progress—manifested through the process of “creative destruction”—serves as a fundamental engine for economic expansion (Aghion et al., 2005; Lokuge et al., 2025). DT fosters an environment conducive to such creative destruction and holds significant potential for enhancing productivity and promoting high-quality economic development (Peralta et al., 2025). This perspective is consistent with the technology–organization–environment framework (Hoque et al., 2025), which highlights how technological innovations interact with organizational and environmental contexts to shape performance outcomes.
From a political science perspective, NQPF not only represents the continuation and evolution of the Marxist theory of productive forces, but also reflects the deep integration of data as a fundamental production element (Fu & Zhao, 2025). At its core, NQPF emphasizes the optimized combination of laborers, means of production, and objects of labor (Ren et al., 2025). The introduction of this concept enriches the theoretical framework of Marxist productive forces theory and offers both theoretical support and practical guidance for advancing Chinese-style modernization (Yang & Che, 2025). Drawing on Schumpeter’s theory of innovation and endogenous growth, technological progress, operating through the mechanism of “creative destruction,” constitutes a key driver of economic growth (Lokuge et al., 2025). DT fosters an environment conducive to such creative destruction and holds significant potential for enhancing productivity and promoting high-quality economic development (Peralta et al., 2025). Furthermore, the deep integration of DT into the economy and society facilitates the transformation of both production factors and production relations toward digitalization (Chin et al., 2025). This transformation provides high-quality labor and advanced technological means of production, thereby enabling a substantial leap in overall productivity (X. Liu & Liu, 2025). Enterprise-level DT has the potential to cultivate new types of workers, labor resources, and labor objects, thereby empowering the development of NQPF (Yao, 2024). First, digital transformation within enterprises is driven by DT as a core engine, characterized by its strong permeability and spillover effects (Zhou & Hu, 2024). It enables digital elements to penetrate all aspects of production and operations, thereby optimizing the configuration of production factors and improving total factor productivity (Ding et al., 2025). Second, the implementation of digital technologies necessitates a highly skilled workforce, which promotes the optimization of the human capital structure, fosters the emergence of new types of workers, and supports the internal development of NQPF (Fazlollahi & Franke, 2018). Third, DT enhances the efficiency of traditional production inputs through applications such as big data analytics, facilitating the evolution of new types of labor tools and further promoting NQPF development (Zhao et al., 2022). Finally, by driving green innovation, DT boosts enterprise performance while stimulating demand for new energy and advanced materials through market mechanisms, consequently creating new labor objects and strengthening enterprise NQPF development (J. Wang et al., 2025). On this basis, the study advances the following hypothesis:
The Mediating Role of Green Technology Innovation and Management Efficiency
When elaborating on the theoretical connotation of NQPF, President Xi Jinping emphasized that its core characteristic is green development (Yue et al., 2024). Technological innovation acts as the core driving force of economic growth, allowing it to overcome the law of diminishing returns through the promotion and application of green technological innovation and advanced green technologies (Ohikhuare et al., 2025). Firstly, DT offers robust data processing and analytical capabilities, enabling enterprises to make scientific decisions in green technology innovation (Khan et al., 2025). Through big data analytics, enterprises can collect and process vast amounts of environmental data, identifying inefficiencies in resource utilization, energy consumption, and pollution emissions, thereby identifying areas for optimization (Siripi et al., 2024). Data-driven decision support allows enterprises to accurately identify key areas for green technology R&D and allocate resources effectively (Y. Liu et al., 2024). Secondly, DT enhances production processes through real-time monitoring and data feedback, minimizing resource waste and environmental pollution (Heo et al., 2025). Intelligent manufacturing technologies facilitate real-time communication and collaboration between equipment and systems, enabling dynamic adjustments to production parameters to achieve optimal production efficiency with minimal environmental impact (Tan & Yu, 2024). Moreover, simulation and modeling tools supported by DT are essential in the R&D phase of green technologies (J. Li et al., 2024). Using virtual reality (VR) and augmented reality (AR) technologies, enterprises can evaluate the feasibility and environmental performance of new technologies in virtual environments, thereby reducing trial-and-error costs and accelerating the pace of technological innovation. This logic is supported by evidence that “smarter” production processes, underpinned by AI and intelligent manufacturing, are inherently “greener” and lead to superior environmental, social, and governance outcomes (Gao et al., 2025). On this basis, the study advances the following hypothesis:
Enterprise NQPF is defined as the enhanced production capacity achieved through management optimization and process transformation while adhering to the principles of sustainable development (J. Wang et al., 2025). Firstly, DT facilitates the automation and intelligence of management processes (Moderno et al., 2024). For example, Enterprise Resource Planning (ERP) systems integrate multiple modules, including finance, supply chain, manufacturing, and human resources, thereby enabling centralized information management and process automation, which significantly enhances operational efficiency. Customer Relationship Management (CRM) systems analyze customer data, enabling enterprises to formulate precise marketing strategies, thereby enhancing customer satisfaction and loyalty, as well as optimizing resource allocation and market responsiveness. Secondly, the application of DT in management significantly improves efficiency (Amer et al., 2024). For instance, artificial intelligence is used for predictive maintenance, which involves analyzing real-time equipment operation data to identify potential issues in advance, thereby preventing production interruptions and reducing maintenance costs. Moreover, AI optimizes supply chain management through demand forecasting and inventory optimization, thereby reducing inventory costs and ensuring production continuity and flexibility. Furthermore, DT enables real-time monitoring and data feedback through the interconnection of equipment and systems, thereby improving the precision of management (H. Zhang et al., 2024). The application of DT in supply chain management enhances transparency and traceability. Blockchain offers a transparent and tamper-proof platform for recording the entire process of products, spanning raw material procurement, production, transportation, and sale (Njualem, 2022). This transparency not only enhances supply chain efficiency but also strengthens product quality control and compliance management, thereby reducing management risks. On this basis, the study advances the following hypothesis:
The Moderating Role of Market Competition Intensity
In today’s highly competitive business environment, enterprises face sustained pressure to innovate and enhance productivity to maintain their competitive edge. Market competition intensity is a key factor shaping corporate strategic decisions and innovation behaviors, particularly in the context of rapid DT development, where its influence becomes even more pronounced (M.-J. Chen, 1996). While DT is often applied to improve operational efficiency rather than to foster disruptive innovation, its impact on enhancing enterprise NQPF remains limited. However, as competition intensifies, enterprises encounter increasingly complex and diverse challenges, rendering traditional resources and capabilities inadequate to address these uncertainties (Hart & Ahuja, 1996). In such an environment, persistent innovation becomes essential, making digital transformation a critical strategic choice. This logic is underpinned by the “escape-competition effect” from industrial organization theory (Aghion et al., 2005). Intense competition erodes the profits derived from incumbent technologies and products. To escape this competitive pressure and secure transient market power, firms are compelled to engage in more radical, “neck-and-neck” innovation (Y. Chen et al., 2015). Digital technologies, characterized by their disruptive potential, serve as a primary tool for such leapfrogging. Consequently, in highly competitive markets, the marginal returns on investing in DT are higher. Firms are not only pressured to adopt DT for operational efficiency but are also driven to leverage it for creating differentiated products and services to survive and outperform rivals (Nguyen et al., 2025). Therefore, we posit that market competition intensity positively moderates the DT-NQPF relationship: the fiercer the competition, the more firms are forced to rely on the transformative power of DT to drive the innovative and qualitative leaps that constitute NQPF. On this basis, the study advances the following hypothesis:
Research Design
Variable Definition
Explained Variables
The measurement of New Quality Productive Forces (NQPF) is constructed upon its theoretical conception as a qualitative leap in productive elements. We developed a comprehensive indicator system comprising three core dimensions: new-quality laborers (as the dynamic core), new-quality labor objects (as the action target), and new-quality labor materials (as the transformation engine). This framework, informed by Yao et al. (2025) and China’s theoretical discourse, facilitates a scientific assessment of enterprise NQPF development (as shown in Table 1). Subsequently, the entropy weight-TOPSIS method was employed to quantify the NQPF index at the enterprise level.
Measurement Index System of Enterprise NQPF.
Explanatory Variable
Digital Technology (DT): Drawing on the research of Yu et al. (2024) and Luo et al. (2025), this study utilized textual analysis to measures firms’ digital technology. The primary data source is the “Management Discussion and Analysis (MD&A)” section of listed companies’ annual reports, which provides a comprehensive overview of firms’ strategies and practices in digital transformation. To construct the measure, a digital dictionary was first developed by combining keywords from policy documents, the World Intellectual Property Organization’s technical glossary, and existing literature. The dictionary was further expanded using the RoBERTa-wwm-ext deep learning model (Cui et al., 2021), retaining terms with a similarity score above 0.8. The final dictionary includes 97 keywords, covering four main themes: digital technology application, internet business models, intelligent manufacturing, and advanced information systems. Based on this dictionary, digital-related sentences were extracted from annual reports, and the ratio of their word count to the total word count was calculated. This proportion, multiplied by 1,000, serves as the core index of digital technology adoption at the firm-year level. To ensure robustness, an alternative measure was also employed, defined as the logarithm of the total frequency of digital-related terms. The consistency of empirical results across these alternative measures confirms the reliability and explanatory power of the constructed indicator.
Mediating Variables
This study chose green technology innovation (Green) and management efficiency (Me) as the mediating variables. The measurement of Green was operationalized as the natural logarithm of one plus the total count of green utility model patents and green invention patents, following the approach of Y. Wang et al. (2025). For management efficiency (Me), this study adopted the inventory turnover ratio as a proxy, a metric informed by the work of Qiao and Chen (2025). This ratio is calculated as the cost of goods sold divided by the average inventory, where a higher value signifies superior management efficiency.
Moderating Variables
The intensity of market competition was captured by the Herfindahl-Hirschman Index (HHI), which serves as the moderating variable in this study, following Hu et al. (2025). This index is computed as the sum of the squares of all individual firms’ revenue shares within their industry. A lower HHI denotes a more competitive and fragmented market structure, whereas a higher value signals greater concentration and reduced competition. To facilitate a more intuitive interpretation, we employed the reciprocal of the HHI, thereby ensuring that an elevated value on this transformed scale aligns directly with heightened competitive intensity.
Control Variables
Drawing on the studies of Qi et al. (2024) and Wei et al. (2024), this study selected a comprehensive set of indicators that may influence the development level of NQPF in listed companies. These indicators encompass enterprise value (Tq), asset-liability ratio (Lev), firm size (Size), firm age (Age), cash holdings (Cash), board size (Board), proportion of independent directors (Ddbl), and CEO duality (Dual). Table 2 provides the complete operational definitions and measurement methodologies for these variables.
Explanations of Variables.
Data Sources
This study focused on the analysis of Chinese A-share listed companies spanning the period from 2011 to 2023. Before 2011, the adoption of digital technologies by enterprises was relatively limited. However, the rapid development of mobile internet in 2011 significantly accelerated the digital transformation of enterprises, with the ap-plication of digital technologies expanding in both breadth and depth. The research sample was rigorously screened, excluding companies with missing key financial indicators, incomplete data, financial firms, and those exhibiting abnormal operations, such as ST or *ST status. Consequently, 25,828 valid observations were retained for analysis. The primary data sources include the CSMAR and Wind databases, which provide comprehensive data on assets and liabilities, profits, cash flows, stock capital markets, and enterprise characteristics. These comprehensive and systematic datasets offer robust support for the study, facilitating a thorough analysis of the impact of DT on enterprise NQPF and ensuring the reliability and scientific rigor of the research findings. To mitigate the influence of outliers, winsorization was applied, restricting all variable values to the range between the 1st and 99th percentiles.
Model Specification
To test Hypothesis H1, a linear regression model controlling for province, industry, and year fixed effects is constructed. This model is represented by Equation (1).
To test Hypotheses H2 and H3, which examine whether green technology innovation (Green) and management efficiency (Me) mediate the impact of DT on enterprise NQPF, this study constructs Equations (2) and (3) as the basis for the mediation analysis:
To test Hypothesis H4 and examine whether market competition intensity (HHI) moderates the effect of DT on enterprises’ NQPF, Equation (4) is constructed:
where the subscripts a, j, k, and t represent firm, province, industry, and year, respectively; NQPF denotes the enterprise’s level of new quality productive forces, while DT measures the degree of digital technology application. The variables Green and Me serve as mediators, reflecting green technology innovation and management efficiency, respectively. HHI captures market competition intensity and functions as the moderating variable, and DT×HHI is the interaction term combining the core explanatory variable with the moderator. Controls comprises the set of control variables; σ, γ, and δ represents province, industry and year fixed effects, respectively.ε denotes the random error term.
Analysis of Empirical Results
Descriptive Statistics
Table 3 presents the descriptive statistics for all variables. The mean value of NQPF is 5.072 and the median is 4.693, suggesting a right-skewed distribution. This indicates that while some enterprises have achieved significant improvements in new quality productive forces, the majority of firms remain below the overall average level. Meanwhile, the substantial variation in NQPF (ranging from 0.594 to 14.920) highlights pronounced heterogeneity in firms’ productivity transformation and upgrading progress. For the core explanatory variable DT, the mean and median are 1.235 and 0.694, respectively. The noticeable difference between these two values implies that digital transformation is relatively weak for most enterprises, and only a limited number of firms have made substantial investments in digital technology applications. This distribution pattern also suggests a broad space for further digital upgrading in China’s manufacturing sector. Overall, these results reveal that both NQPF and DT are characterized by substantial inter-firm heterogeneity, reinforcing the necessity of further econometric analysis to explore whether and how digital technology affects the enhancement of new quality productive forces.
Descriptive Statistics of Variables.
Benchmark Regression Results
Based on the regression model previously established in Equation (1), this study assesses the influence of DT on enterprises’ NQPF. Table 4 summarizes the baseline estimation outcomes. Column (1) presents the initial association between DT and NQPF without additional controls. Columns (2) and (3) introduce fixed effects and control variables sequentially. Column (4) displays the fully specified model, incorporating all control variables along with industry and year fixed effects. Across each specification, the estimated coefficient associated with DT is positive and reaches statistical significance at the 1% level. This pattern confirms that digital transformation plays a substantial role in elevating enterprise NQPF, providing consistent support for Hypothesis H1.
Benchmark Regression Results.
Note. ** and *** respectively denote the p < .05 and p < .01; the value in brackets is the robust standard error.
Robustness Tests
Adjusting the Research Sample
To mitigate potential disruptions caused by the COVID-19 pandemic—which introduced systemic economic risks and affected both supply and demand conditions from early 2020 onward—we first re-estimated the regression using a truncated sample that excludes all observations from 2020 and later. The outcome, displayed in column (1) of Table 5, remains consistent with our main specification. Second, firms headquartered in municipalities directly administered by the central government were removed from the sample. These regions often implement distinctive policy measures, which may disproportionately influence local firms and introduce estimation bias. After omitting these entities, the model was re-run. Results reported in column (2) of Table 5. Notably, even after these adjustments to the study period and sample composition, the coefficient of DT with respect to NQPF remained positive and significant at the 1% level. These findings underscore the consistency and reliability of our core results.
Robustness Tests.
Note. ** and *** respectively denote the p < .05 and p < .01; the value in brackets is the robust standard error.
Alternative Measures of DT and NQPF
First, we replaced the core explanatory variable measurement indicators. Different DT measurement indicators may affect their effectiveness in promoting the NQPF of enterprises. This study adopted the method of Kong et al. (2024) and employs the logarithm of the total number of digital-related terms as a proxy independent variable to re-examine its impact on enterprise NQPF. The results are shown in column (3) of Table 5. Second, the calculation method for the explained variable was adjusted. Specifically, the measure of enterprises’ NQPF was shifted from the entropy-weighted TOPSIS approach to principal component analysis, and the model was re-estimated using this revised measure; the results are reported in column (4) of Table 5. Whether the change concerned the explanatory or the dependent variable, the estimated coefficients of DT consistently remained positive and statistically significant at the 1% level, suggesting a strong and favorable influence of DT on NQPF. These outcomes are fully consistent with the baseline estimates, further attesting to the robustness of the documented relationship between DT and NQPF.
Interaction Fixed Effects
To better control for unobserved factors that may vary across regions and sectors over time, this study augments the baseline specification with a set of high-dimensional fixed effects. In addition to controlling for industry, region, and year separately, we further include province-year and industry-year interaction terms. This approach helps absorb time-varying shocks at the provincial and industrial levels, such as shifts in regional policy or sector-specific technological trends. As shown in column (5) of Table 5, the estimated coefficient of digital transformation remains positive and statistically significant at the 1% level even under this more stringent set of controls. This reaffirms that DT exerts a substantial and robust positive influence on the development of enterprise NQPF, even after accounting for a richer structure of unobserved heterogeneity.
Change the Clustering Hierarchy
To mitigate potential issues of serial correlation and heteroscedasticity, this study relaxed conventional regression assumptions by clustering standard errors at a higher aggregation level. Specifically, robustness checks are conducted using industry-level cluster-robust standard errors, with the results reported in column (6) of Table 6. Considering that local governments in China often exert substantial influence over regional economic and policy environments (C. Xu, 2011), error terms may exhibit correlation at the region-industry level. To address this, the bidirectional clustering approach proposed by Cameron et al. (2011) is applied to simultaneously cluster standard errors by both industry and region. The corresponding results, shown in column (7) of Table 7, confirm that the key findings remain robust across different clustering specifications.
Endogeneity Test.
Note. *, ** and *** respectively denote the p < 0.1, p < 0.5 and p < 0.01; the value in brackets is the robust standard error.
Mediating Effects Analysis.
Note. *** denote p < .01; the value in brackets is the robust standard error.
Placebo Test
To mitigate potential confounding from time trends or other coincidental factors in the observed relationship between DT and NQPF, we implemented a placebo test using randomly generated DT indicators to verify the robustness of our core results. The specific operational steps are as follows: first, construct pseudo-firm DT dummy variables through random assignment; second, replace the DT variables used in the preceding text with the pseudo-firm DT dummy variables, and substitute the sample into model (1) for regression testing; third, repeat the aforementioned process 1,000 times; the placebo test results are shown in Figure 1. From the estimated coefficients, a symmetrical distribution structure centered on 0 was observed. The mean of the estimated coefficients differs significantly from the coefficient value of 0.175 in Table 3. The distribution of the p-values shows that most random sample regression coefficients have p-values greater than .1, meaning that there is no significant correlation. The placebo test results indicate that the relationship between DT and new productive forces in the main test is indeed valid, and to a certain extent, the influence of time variables and other unknown random factors is ruled out.
Placebo test.
Endogeneity Test
Instrument Variable Method
The nexus between DT and NQPF may be subject to endogeneity issues. While DT can facilitate the advancement of NQPF, firms with inherently superior productivity may possess greater resource availability and stronger strategic motivations, thereby increasing their likelihood of adopting digital technologies—creating potential reverse causality. In addition, NQPF is shaped by various complex and potentially unobservable determinants, which could lead to omitted variable bias and complicate causal inference. To mitigate these endogeneity concerns, this study employs a two-stage least squares (2SLS) estimation with instrumental variables. Drawing on the methodology of Cirillo et al. (2024), the first instrumental variable (IV1) is the lagged value of the enterprise’s DT level, which serves as a proxy for the current DT. In addition, following Mi et al. (2024), the second instrumental variable (IV2) is constructed as the industry-province-level average of DT, capturing broader regional and sectoral trends. These instruments are chosen as they satisfy the exclusion restriction by capturing the external digital climate, which affects firm-level NQPF only through its impact on the firm’s own DT investment and adoption decisions. Table 6 reports the 2SLS estimation outcomes. Columns (1) to (2) display results using IV1, whereas columns (3) to (4) present those based on IV2. The estimates show that both instrumental variables yield significantly positive coefficients, demonstrating their strong explanatory power for firms’ digital adoption. Furthermore, the DT coefficient remains significantly positive across all specifications, corroborating that digital transformation markedly promotes enterprises’ NQPF.
Moreover, the Kleibergen-Paap rk LM statistic is significant at the 1% level, rejecting the null of under-identification and confirming the relevance of the instrumental variables. The Kleibergen-Paap rk Wald F statistic is well above the Stock-Yogo 10% critical value of 16.38, indicating that weak instrument concerns are negligible. Moreover, the endogeneity test is significant at the 1% level, rejecting the exogeneity of the explanatory variables and verifying the presence of endogeneity in the model. Collectively, these diagnostics demonstrate that the instruments used are both valid and reliable. After correcting for endogeneity, the DT coefficient remains significantly positive, reinforcing the robustness of the baseline results and offering strong empirical support for Hypothesis H1, further highlighting DT’s pivotal role in advancing enterprise NQPF.
Multiple-Period Difference-in-Differences (DID) Model
To further address potential endogeneity issues and enhance the robustness of the empirical analysis, this study exploits China’s Big Data Comprehensive Pilot Zone policy as a quasi-natural experiment. The difference-in-differences (DID) framework is applied to identify the causal influence of DT on enterprises’ NQPF. Guizhou Province initiated the nation’s first Big Data Pilot Zone in 2015, after which several additional provinces were approved, resulting in the establishment of ten national-level pilot zones by 2019. Leveraging this staggered policy rollout, the Big Data initiative is treated as an exogenous policy intervention, forming the basis for constructing a multi-period DID model. The model can be specified as follows:
In this model, the variable treat × post represents the interaction term for the Big Data pilot policy. Specifically, treat is a binary indicator that equals 1 if the enterprise is located in a city designated as a Big Data Comprehensive Pilot Zone, and 0 otherwise. The variable post denotes the policy implementation period, taking the value of 1 for the year when the city officially launched the pilot and for all subsequent years, and 0 for the pre-policy period. All remaining control variables follow the definitions used in the baseline regression model. As reported in column (5) of Table 6, the estimated coefficient for the Big Data policy variable is significantly positive at the 1% level, suggesting that the introduction of this national pilot policy effectively promoted the enhancement of enterprises’ new quality productive forces (NQPF), thereby providing further empirical support for Hypothesis H1.
When applying the multi-period difference-in-differences (DID) approach, it is essential to verify that the treatment and control groups exhibit comparable trends in the outcome variable prior to the policy intervention. In other words, before the launch of the Big Data pilot, enterprises in both groups should display parallel movements in NQPF levels over time. Following the methodology of Borusyak et al. (2024), this study performs a parallel trend test using the estimation results presented in column (5) of Table 6, where the pre-policy period serves as the baseline reference. As illustrated in Figure 2, the horizontal axis depicts the relative years surrounding the initiation of the Big Data policy, with “0” marking the year of implementation. Years positioned to the left correspond to the 4 years preceding the policy, while those on the right indicate the subsequent 7 years after the pilot began. The vertical axis reflects the estimated coefficients of the policy effect. Coefficient values above the zero baseline suggest statistically significant deviations following the policy intervention. The overall pattern demonstrates that the treated and control enterprises followed a consistent trajectory before the policy took effect, thereby confirming the validity of the parallel-trend assumption required for the DID model.
Result of parallel-trend test.
PSM Testing
To mitigate potential endogeneity and sample selection bias, this study further employs the Propensity Score Matching (PSM) approach. Firms are divided into a treatment group and a control group according to the industry-specific median value of DT adoption for each year. Enterprises with DT levels equal to or above the industry median are assigned to the treatment group, while those below the median form the control group. Using firm-level covariates such as enterprise value, leverage ratio, firm size, firm age, cash holdings, board size, and the proportion of independent directors, the matching is conducted based on the nearest neighbor algorithm with replacement (1:1 matching). As shown in Figure 3, the matched samples are symmetrically distributed around zero, indicating that the balancing property is satisfied and the matching quality is acceptable. After matching, the adjusted sample is used for regression analysis. The results reported in column (6) of Table 5 show that the estimated coefficient of DT remains positive and statistically significant at the 1% level, indicating that digital technology effectively promotes enterprises’ NQPF, thereby providing further empirical support for Hypothesis H1.
Propensity Score Matching (PSM).
Further Analyses
Analysis of Mediating Effects
Mediating Effects of Green Technology Innovation
In the context of the digital economy, digital technology does not only directly strengthen the productive capacity of enterprises; more critically, it acts as a catalyst for corporate high-quality development by fostering green technology innovation—an essential mediating pathway. Serving as a strategic pivot in the digital transformation process, green technology innovation effectively bridges DT and NQPF by enabling an efficient transmission of value. As reported in column (1) of Table 7, where green technology innovation is employed as a mediating variable, the net regression coefficient of DT is 0.129, which is statistically significant at the 1% level. This provides empirical support that DT actively facilitates green technology innovation, thereby confirming the presence of a mediation effect. Further validation comes from the Sobel and Bootstrap test, which yields a Z-value of 7.693 (p < .01), and the 95% Bootstrap confidence interval that does not include zero. Together, these results substantiate that green technology innovation serves as a significant mediator in the relationship between DT and enterprises’ NQPF, thus supporting Hypothesis H2. This mediating mechanism can be attributed to the fact that DT establishes an enabling environment for green technology innovation through enhanced information processing, optimized resource allocation, and accelerated R&D cycles. In turn, green technology innovation translates the facilitative role of DT into tangible productivity improvements by lowering energy consumption and reducing pollution emissions, thereby ultimately advancing the development of corporate NQPF.
Mediating Effects of Management Efficiency
To reinforce new productive capacity, enterprises should intensify investment in digital transformation (DT), promote technological upgrading, and optimize operational and inventory control. A higher inventory turnover ratio typically signifies more effective utilization of inventory resources and greater managerial efficacy, which collectively support the development of new quality productive forces. Following the implementation of digital transformation, firms generally achieve faster inventory circulation and notable gains in operational performance. As shown in column (2) of Table 7, which reports regression results with the inventory turnover ratio as the mediating variable, the coefficient of DT is significantly positive at the 1% level. This implies that DT contributes to improved inventory turnover, thereby elevating management efficiency and facilitating the enhancement of new quality productivity at the enterprise level. The mediating effect was further corroborated through both the Sobel and Bootstrap test. The Sobel test produced a Z-value of −2.734 (p < .01), and the 95% confidence interval derived from Bootstrap sampling excluded zero, confirming a statistically significant mediation pathway. These findings validate that the inventory turnover ratio serves as a mediator in the relationship between DT and NQPF, supporting Hypothesis H3. The mechanism underlying this effect can be attributed to the capacity of DT to refine inventory management via real-time data monitoring, automated replenishment systems, and intelligent forecasting capabilities. Such improvements in inventory turnover help lower holding costs, enhance the efficiency of capital use, and strengthen operational flexibility, thus collectively promoting the growth of new quality productive forces.
Analysis of Moderating Effects
Building on the theoretical framework established earlier, this study hypothesizes that the intensity of market competition strengthens the positive association between DT and enterprises’ NQPF. An empirical test was conducted by estimating Model (4), with the complete findings displayed in Table 8. Column (1) provides the baseline estimate of DT’s influence on NQPF in the absence of the moderating variable. Column (2) introduces market competition intensity into the analysis. The regression coefficient for market competition intensity is 0.173, statistically significant at the 1% level, implying that competitive pressure independently contributes to NQPF development. Furthermore, the interaction term between market competition intensity and DT shows a coefficient of 0.176, also significant at the 1% level. This result demonstrates that under equivalent degrees of digital transformation, a more competitive market environment significantly amplifies the positive effect of DT on NQPF. Stated differently, market competition intensity functions as a meaningful moderator in the relationship between digital transformation and enterprise productivity enhancement, thereby providing empirical confirmation for Hypothesis H4.
Moderating Effects Analysis.
Note. *** denote p < .01; the value in brackets is the robust standard error.
Heterogeneity Analysis
Ownership Heterogeneity
The influence of digital transformation is likely mediated by a firm’s ownership type, largely because of variations in resource availability and institutional environments. Theoretical reasoning suggests that State-Owned Enterprises (SOEs) are uniquely positioned to leverage DT for NQPF development. First, SOEs typically possess superior resource endowments, including easier access to capital and key infrastructure, which are critical for funding large-scale digital transformation projects. Second, they operate under stronger institutional and policy mandates to lead in national strategic priorities, which now explicitly include the development of NQPF and green innovation. Given this synergy between resource capacity and policy orientation, we propose that the favorable impact of DT on NQPF will be stronger within SOEs than in non-SOEs. As illustrated in Figure 4, DT coefficients are positive and statistically significant for both ownership categories. Notably, the estimated effect for state-owned enterprises markedly exceeds that of non-state-owned entities. A Bootstrap test with 1,000 replications yields an intergroup difference with a p-value of .000, validating a statistically meaningful divergence in how DT influences NQPF across ownership types. In summary, while digital transformation strengthens the NQPF of all firms, its benefits are more substantial in the state-owned sector.
Heterogeneity analysis results.
Environmental Pollution Heterogeneity
The inherently green nature of NQPF creates a theoretical expectation of heterogeneity based on a firm’s environmental footprint. We posit that the transformative potential of DT is structurally constrained in heavily polluting enterprises. Such firms often face a “lock-in effect” or path dependency rooted in high-emission production technologies and legacy assets. This technological trajectory fundamentally conflicts with the core objective of NQPF, which is to decouple economic growth from environmental damage. Consequently, even if DT offers optimization potential, its capacity to drive a fundamental shift towards new quality productivity is likely inhibited in heavy polluters. We therefore expect a weaker or insignificant effect of DT on NQPF for heavily polluting firms. Enterprises were classified into heavily polluting and non-heavily polluting categories according to the Industry Classification Management Catalogue for Environmental Protection Verification of Listed Companies issued by China’s Ministry of Environmental Protection in 2008. The regression results, reported in Figure 4, indicate that the DT coefficient is insignificant for heavily polluting enterprises but significant for their non-heavily polluting counterparts. This may be because new-type productive forces are inherently green, whereas heavily polluting enterprises, characterized by high carbon emissions, are misaligned with green development, exerting a natural inhibitory effect. Consequently, the polluting nature of these enterprises diminishes the capacity of DT to enhance NQPF.
Technological Heterogeneity
The efficacy of DT may be amplified in contexts where technology is a central production factor. A priori, there are compelling reasons to expect a stronger DT-NQPF link in high-tech enterprises. The very definition of these firms implies a higher inherent compatibility and complementarity between their core knowledge base, R&D processes, and digital technologies. This synergy allows for a deeper integration of DT, enabling not only incremental efficiency gains but also more radical innovation in products and processes—the hallmark of NQPF. Furthermore, high-tech firms’ reliance on intellectual property and rapid innovation aligns perfectly with DT’s capability to manage knowledge and accelerate iteration. Thus, we hypothesize that the positive impact of DT on NQPF will be significantly stronger for high-tech firms. As reported in Figure 4, the DT coefficients are significantly positive in both groups; however, the effect is notably larger for high-tech firms. The empirical p-value for the coefficient difference, obtained from 1,000 Bootstrap replications, is 0.000, confirming a statistically significant disparity between the two groups. These findings indicate that high-tech attributes amplify the positive impact of DT on enterprises’ NQPF. A plausible explanation is that high-tech enterprises prioritize cutting-edge technological R&D and allocate greater resources—such as human and financial capital—during operations, thereby enabling their digital transformation to exert a stronger catalytic influence on new quality productivity.
Regional Heterogeneity
President Xi Jinping has underscored that the cultivation of NQPF should align with local realities, reflecting the distinct conditions of each region. Given the heterogeneous foundations across regions, policy measures should be adapted to their specific circumstances. In regions with well-established digital infrastructure, firms are more capable of leveraging digital transformation to optimize managerial processes and accelerate environmentally-friendly innovation. It follows that the contribution of DT to NQPF is likely to be more substantial in digitally advanced areas. To examine regional differences, companies in the sample were classified into eastern, central, and western zones according to their registered locations. Regression estimates, presented in Figure 4, demonstrate that DT exerts a significant positive effect on NQPF among firms based in the eastern region. In contrast, the relationship is not statistically significant for those in the central and western parts of China. This variation may be attributed to the eastern region’ s superior digital facilities, deeper reserves of specialized digital human capital, and more dynamic competitive landscape—conditions that collectively support more comprehensive digital adoption and amplify DT’ s role in advancing NQPF.
Industry Heterogeneity
The development of NQPF is highly dependent on innovation and the ability to reconfigure factors of production. Technology-intensive enterprises, with their strong R&D foundations, high-quality talent reserves, and high compatibility between technology and data elements, can more effectively integrate DT to unlock its potential in optimizing processes and driving innovation, thereby significantly enhancing new quality productive capacity. In contrast, asset-intensive enterprises may be constrained by high transformation costs and the rigidity of existing assets, while labor-intensive enterprises face challenges in skill matching and management model transformation, resulting in relatively limited or insignificant enabling effects of DT. Following the 2012 industry categorization framework issued by the China Securities Regulatory Commission, firms in the sample were segmented into labor-intensive, capital-intensive, and technology-intensive groups. Regression estimates, illustrated in Figure 4, reveal statistically negligible effects of DT in both labor- and capital-intensive sectors. In contrast, technology-intensive industries exhibit a strongly positive and statistically significant coefficient at the 1% level, with its magnitude considerably surpassing those in the other segments. This divergence can be ascribed to the inherent strengths of technology-intensive companies in human capital and technological endowments, which support more effective fusion of digital tools with production factors. Such integration promotes dual innovation in technology and management, equipping firms with enhanced productive capabilities and technical competencies, and in turn accelerating the evolution of NQPF.
Conclusions
Main Conclusions and Insights
Drawing on panel data from Chinese A-share listed companies spanning 2011 to 2023, this research systematically examines how digital transformation influences enterprises’ new quality productive forces. The study further analyzes heterogeneous effects across organizational and regional contexts, while investigating the mediating functions of green technology innovation and management efficiency, together with the moderating influence of market competition intensity. Key findings are summarized as follows: First, digital transformation exerts a statistically significant and positive effect on enterprise NQPF. This result holds across a series of robustness checks, confirming the stability of the core relationship. Second, the analysis identifies two central transmission channels: green technology innovation and management efficiency. Both serve as meaningful mediators in the DT–NQPF link. At the same time, market competition intensity strengthens the relationship, acting as a significant positive moderator. Third, the productivity-enhancing effect of DT is more strongly observed among firms in the eastern region, state-owned enterprises, and technology-intensive industries.
It should be emphasized that these findings are contextualized within China’s distinctive institutional setting, which includes its unique digital policy system, governance model, and phase of economic development. The specific ways in which digital technologies shape new quality productive forces may differ across countries with varying market institutions, regulatory systems, and technological environments.
Based on the empirical results, this study puts forward the following policy recommendations to support the digital transition of Chinese firms and stimulate high-quality growth:
First, government support should be tailored to the heterogeneous needs of different enterprises. Instead of one-size-fits-all policies, we recommend: For Small and Medium-sized Enterprises (SMEs), governments and industry associations should develop and promote “lightweight” digital transformation toolkits and affordable cloud-based solutions to lower the entry barrier for digital adoption. For Heavily Polluting Enterprises, environmental subsidies and compliance approvals should be explicitly tied to the integrated adoption of digital and green technologies, creating a strong incentive for synergistic transformation.
Second, regional disparities must be addressed through targeted fiscal mechanisms. We propose the establishment of a “Regional Digital Transformation Pilot Fund” specifically for the central and western regions. This fund should prioritize investments in localized digital infrastructure and provide matching grants to attract and cultivate digital talent in these areas, thereby narrowing the digital divide identified in our heterogeneity analysis.
Third, policymakers should design tailored assistance programs that account for firm size and sectoral attributes. SMEs should gain improved access to specialized digital consultation and financial subsidies. High-technology firms ought to be incentivized to bolster R&D in core technologies, thereby elevating their new quality productive forces. State-owned enterprises can serve as stabilizing anchors by facilitating digital upgrading across supply chains and fostering the robust growth of private firms—strengthening overall productive efficiency. In parallel, public initiatives should advance the deployment and use of digital infrastructure in both eastern and central-western regions, aiming to narrow interregional digital disparities and raise nationwide digitalization.
Limitations and Future Prospects
This study has several limitations that warrant attention. First, the exclusive focus on Chinese A-share listed companies, while ensuring data standardization, limits the generalizability of our findings to other contexts. The unique characteristics of China’s institutional environment and the specific features of A-share listed companies represent important boundary conditions for our results. Future research could test the generalizability of our findings in different institutional settings and with diverse firm types. Second, the measurement methods for core variables, particularly “new quality productive forces” and “digital technology,” require further refinement and validation in different contexts. Future research could develop more comprehensive evaluation systems that account for cross-country institutional differences. Third, this study does not fully account for how different national innovation systems and digital policy frameworks might shape the relationship between digital transformation and productivity development. Comparative studies across different countries and institutional environments would be valuable for understanding the boundary conditions of our findings.
Footnotes
Acknowledgements
We will be grateful to the anonymous reviewers who will comment on this manuscript.
Ethical Considerations
All procedures performed in this study were in accordance with the ethical standards of the university.
Funding
The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by National Social Science Fund of China (Grant No. 19ZDA048), Humanities and Social Sciences Research Projects in Jiangxi Universities (Grant No. JJ24216), Science and Technology Project Founded by the Education Department of Jiangxi Province (Grant Nos. GJJ2202813, GJJ2502709), Social Science Planning Project of Nanchang City (Grant No. YJ202406, YJ202409), Jiangxi Institute of Fashion Technology Special Funding Project (Grant No. JFZX-202502).
Declaration of Conflicting Interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Data Availability Statement
All data used in this study have been deposited in the https://github.com/tujian123/Dig. The micro-level enterprises data in the original dataset are sourced from https://data.csmar.com. The city-level data are obtained from 
. All these data are publicly accessible.