Introduction

The concept of new quality productivity represents a qualitative leap from traditional productivity1, focusing on the enhanced combination of laborers, labor materials, labor objects, and their optimization. It signifies a new era of high efficiency, high quality, and sustainable productivity. High-quality development is the primary task in the comprehensive construction of a modern socialist country. Closely related to new quality productivity is total factor productivity. However, new quality productivity emphasizes the realization of high-tech productivity achieved by transcending traditional modes of economic growth. It focuses on qualitative leaps in productivity through revolutionary technological breakthroughs, innovative allocation of production factors, and deep industrial transformation and upgrading. In contrast, total factor productivity emphasizes the increase in output driven by technological progress beyond traditional production factors, with a focus on improving the overall efficiency of these factors. It is achieved through technological advancements, organizational innovations, and management improvements. The enhancement of total factor productivity lays a solid foundation for developing new quality productivity in technological innovation and industrial upgrading.

Although new quality productivity originates in China, its implications extend beyond national borders, holding profound significance for the global community. The development of new quality productivity is expected to drive a reconfiguration of global factors of production, altering their allocation structures. In particular, data as a production factor will become a more dynamic and cross-border element, reshaping allocation methods and fostering the development of new industries and models, thereby transforming the future of the international division of labor. As the international division of labor evolves, corresponding global institutions are also likely to undergo adjustments, with the establishment of new rules, regulations, management frameworks, and standards. Its transformation will necessitate a recalibration of nations’ rights, responsibilities, and interests. Advancing new quality productivity represents a forward-looking strategic position for China and aligns with the shared development aspirations of the global community. However, China faces numerous challenges and obstacles in this endeavor, including insufficient technological innovation, a shortage of digital talent, and the substantial financial investment required. Addressing these pressing challenges in developing new quality productivity has become a critical issue that demands urgent solutions.

Correspondingly, the construction of digital infrastructure itself represents advanced productivity. In terms of enterprise efficiency, digital infrastructure facilitates more efficient and convenient information transfer, with powerful data processing capabilities enabling precise analysis of consumer preferences, optimizing resource allocation, reducing transaction costs, and enhancing operational efficiency. From the perspective of industrial development, digital infrastructure drives the digital transformation of traditional industries, fostering inter-industry collaboration through information sharing across industrial chains. In the realm of innovation, digital infrastructure dismantles information barriers, promotes the flow and sharing of innovative resources, and accelerates the application and dissemination of innovative outcomes in the market. Amid the current wave of technological revolution and digital transformation, the rapid advancement of digital technologies has made digital infrastructure a fertile ground for nurturing new quality productivity. This infrastructure enhances the skills of laborers, facilitates the transformation of unconventional cognitive tasks into conventional ones, and reduces the difficulty of skill acquisition. Furthermore, digital infrastructure endows labor materials with new characteristics, generating positive network externalities and lowering learning costs. It also alters the fundamental logic of labor objects, making technological innovation more flexible and transforming the spatiotemporal structure of innovation. China has prioritized deepening the innovative development of the digital economy and proposed a proactive approach to digital infrastructure construction to accelerate the development of new quality productivity. Digital infrastructure encompasses both the digitization of traditional infrastructure and the creation of intelligent infrastructure that arises alongside advancements in digital technology. It provides foundational support and universal technological assistance for developing new quality productivity, becoming a crucial engine for enabling this transformation and injecting strong momentum into its advancement. Digital infrastructure evolution drives a qualitative change in traditional productivity, making it an inevitable trend to develop new quality productivity in the context of high-quality growth.

Further analysis is essential given the close relationship between digital infrastructure construction and new quality productivity. Can digital infrastructure empower the development of new quality productivity? If so, what pathways facilitate this empowerment? Considering spatial correlations, does local digital infrastructure have spillover effects on neighboring areas’ new quality productivity? Addressing these questions carries significant theoretical value and practical implications for enhancing digital infrastructure to achieve new quality productivity development. Therefore, this study employs a static panel model to empirically examine the impact of digital infrastructure construction on new quality productivity, using a sample of 276 prefecture-level cities in China from 2011 to 2020. A spatial Durbin model will also be constructed to investigate the spatial spillover effects generated by digital infrastructure construction.

Related work

Three key representative works are closely related to the theme of this study. The first addresses the macroeconomic effects of digital infrastructure development; the second examines its environmental governance effects; and the third evaluates the policy effects of digital infrastructure initiatives.

Regarding the macroeconomic effects of digital infrastructure, the literature can be categorized into two distinct viewpoints based on varying research conclusions. The first category posits that the construction of digital infrastructure facilitates economic growth and enhances productivity. Studies in this category argue that digital infrastructure effectively harnesses the potential of broadband technology2, mitigates information asymmetry, and achieves optimal resource allocation. The application of digital technologies is said to drive improvements in innovation efficiency3, consequently boosting corporate productivity and promoting economic growth4,5. Conversely, the second category suggests that digital infrastructure may hinder economic growth and productivity. This perspective posits that over-reliance on digital technologies can stymie economic development in a region6, particularly when the level of digital economic development is mismatched with local infrastructure and economic capabilities, leading to detrimental effects on productivity and overall economic progress.

In terms of environmental governance effects, the literature is again divided into two main viewpoints. The first group of studies asserts that digital infrastructure can reduce carbon emissions and optimize ecological quality. They argue that the development of the digital economy fosters a green transformation in residents’ lifestyles and reduces energy consumption in enterprises7,8,9,10, while enhancing government oversight of environmental quality. The opposing viewpoint contends that digital infrastructure development may further degrade environmental quality. Proponents of this view argue that the rapid advancement of electronic products associated with the digital economy entails significant energy consumption, leading to adverse environmental impacts11,12,13,14,15,16.

Recent literature has treated the “Broadband China” pilot policy as a quasi-natural experiment concerning the evaluation of policy effects from digital infrastructure development. Macroeconomic studies indicate that this policy can enhance technological innovation capabilities17,18, promote green technology innovation19,20, and contribute to high-quality economic development as well as reductions in air pollution21,22,23,24,25. At the microeconomic level, existing studies have shown that the “Broadband China” pilot policy can improve enterprise productivity and facilitate digital transformation in businesses26,27.

A comprehensive review of the existing literature reveals several key insights: First, domestic and international scholars have concentrated their attention on the macroeconomic and environmental governance effects of digital infrastructure, with divergent conclusions that underscore the multifaceted impact of digital infrastructure on economic development. Furthermore, while the concept of new productive forces emerges as a novel perspective, few studies have examined the impact of digital infrastructure on these forces, especially considering its role as a crucial engine for invigorating the digital economy. Second, while significant attention has been given to evaluating the “Broadband China” pilot policy, there remains a notable gap in direct investigations of the effects of digital infrastructure on new productive forces, thereby presenting an opportunity for further inquiry.

The marginal contributions of this study are primarily reflected in the three aspects: First, it establishes a logical framework connecting digital infrastructure development with new productive forces, providing a fresh perspective for enhancing the development of new productive forces and offering theoretical justification for advancing digital infrastructure in the context of high-quality growth. Second, it incorporates technological innovation capacity, resource allocation efficiency, and industrial structure upgrading into the framework to uncover their significant roles in empowering new productive forces through digital infrastructure, effectively unraveling the complexities behind this relationship. Third, it examines the spatial spillover effects of digital infrastructure on developing new productive forces in surrounding areas, providing empirical evidence for fostering positive interactions between central cities and their neighboring regions, thus promoting the development of new productive forces.

Theoretical analysis and research hypotheses

Direct effects of digital infrastructure development on new quality productivity

The development of digital infrastructure empowers the advancement of new quality productivity in technology. On the one hand, the enhancement of digital infrastructure facilitates the dissemination of information among enterprises, industries, and municipalities, fostering communication and collaboration across these entities. This creates an effect of information integration and sharing28,29, igniting innovative technologies through the collision of ideas and effectively promoting the growth of new quality productivity in technology. On the other hand, a robust digital infrastructure allows parties with informational advantages to access green innovation-related information more quickly and efficiently. It also helps those with informational disadvantages to identify adequate details about innovation, thereby accelerating the flow of technological innovation factors and optimizing resource allocation19,30,31,32. It minimizes resource waste and inefficiencies in resource matching, further empowering the development of new quality productivity in technology.

Digital infrastructure construction also empowers the development of green productive forces. Achieving advancements in green productive forces requires the reintegration and innovation of knowledge across various fields, including production technology, emission reduction technology, and pollution control technology. Enhanced digital infrastructure can break down geographical limitations on knowledge dissemination, allowing enterprises to integrate information related to green technologies across a broader spatial and temporal spectrum. It focus on developing sustainable industries effectively promotes the growth of green productive forces. Additionally, regions with well-developed digital infrastructure attract technology, talent, and information, harnessing economies of scale. It reduces the search costs for information related to green production technologies, encouraging enterprises to engage in green technology activities. It can lead to optimized management and production processes, ultimately improving production efficiency and resource utilization33,34,35, thus empowering the development of green productive forces.

Furthermore, the construction of digital infrastructure fosters the development of digital productive forces. Enhanced digital infrastructure has led to the emergence of numerous new industries, including artificial intelligence, big data, and the Internet of Things, which provide the support necessary for a leap in productive capabilities and promote the growth of the digital economy. Additionally, improved digital infrastructure contributes to forming a new digital technology system that drives enterprises toward digital transformation and facilitates the transition of traditional industries to intelligent operations, thereby achieving the digital transformation of the real economy. Therefore, we propose Hypothesis 1:

H1: Digital infrastructure construction promotes the development of new productive forces.

The indirect impact channels of digital infrastructure development on new quality productivity

Digital infrastructure development enhances the capacity for technological innovation, thereby facilitating the growth of new quality productivity. The role of digital infrastructure in promoting technological innovation can be highlighted through two primary aspects. First, technological innovation arises from the combination and collision of existing information. Still, the dissemination of this information is often restricted by geographical distance, limiting spillover effects in more distant regions. With the application of digital infrastructure, temporal and spatial distances are compressed, accelerating the flow of information across more expansive areas. This allows for low-cost replication and spillover of information, fostering the spread of innovative technologies and enhancing technological innovation capacity36. Second, the drive and direction of technological innovation stem from consumer preferences. When businesses seek to understand these preferences, they often face information asymmetries, which can reduce their motivation and misdirect their innovation efforts. Implementing digital infrastructure facilitates precise and effective searching and integration of relevant consumer preference information, thereby lowering search, information, and decision-making costs for enterprises. This, in turn, stimulates the motivation for technological innovation and promotes its enhancement37,38.

The improvement in technological innovation capabilities contributes significantly to developing new quality productivity in cities. The impact of technological innovation on new quality productivity is evident in two main ways: First, technological advancements enhance the quality of the workforce. As scientific and technological developments diversify the channels workers acquire knowledge and skills, learning efficiency improves, thus providing a solid labor foundation for empowering new quality productivity. Second, technological innovation introduces new production factors. Continuous advancements in science and technology give rise to novel production elements that differ from traditional factors such as land, capital, and labor; data emerges as a crucial production factor that permeates every aspect of production, further empowering the development of new quality productivity. Therefore, we propose Hypothesis 2a:

H2a: Digital infrastructure construction enhances technological innovation capabilities, thereby empowering the development of new quality productivity.

The development of digital infrastructure enhances the efficiency of factor allocation, fostering the advancement of new productivity paradigms. The impact of digital infrastructure construction on improving factor allocation efficiency is primarily reflected in two key aspects. First, digital infrastructure provides advanced digital technologies and network platforms that facilitate the dissemination and diffusion of factors, streamline the “information arteries”, and promote the synergistic integration of factors, ultimately enhancing allocation efficiency. Second, with its high penetration and strong synergy, digital infrastructure establishes a highly interconnected network structure characterized by positive externalities and scale effects. The network structure creates favorable conditions for the creation, aggregation, transfer, and application of various factors, reducing search costs and mismatch rates, thereby enabling the precise allocation of resources.

Improved factor allocation efficiency directs high-quality talent, advanced production technologies, and credit resources toward highly efficient enterprises and projects. The more efficient the enterprises and projects, the higher the quality of the production factors they attract. This dynamic incentivizes businesses to adopt energy-saving and environmentally friendly production technologies, reducing undesired outputs. The regional aggregation of high-quality factors further enhances knowledge and technology density, generating scale and knowledge spillover effects39. These effects contribute to the improvement of regional production efficiency and facilitate the development of new productivity paradigms. Therefore, we propose Hypothesis 2b:

H2b: Digital infrastructure construction enhances the development of new productivity forms by improving resource allocation efficiency.

The development of digital infrastructure drives the advancement of novel productive forces by fostering industrial structural upgrading. The role of digital infrastructure in promoting the development of digital finance is primarily reflected in the following three aspects: First, digital infrastructure, which relies on new production factors such as data and features core facility systems like 5G communication networks and the Internet of Things, facilitates the comprehensive transformation of production and management models in traditional industries. It drives the upgrading of entire industrial chains, promoting the evolution of low-value-added industries, such as capital-intensive and labor-intensive sectors, toward high-value-added, knowledge-intensive, and technology-intensive industries. Additionally, it spurs the emergence of new industries and business models, further advancing high-tech industries and empowering industrial structural upgrading. Second, the application of digital infrastructure breaks down barriers between industries, fostering cross-sector integration and enabling the rise of emerging industries such as big data and artificial intelligence. Third, digital infrastructure construction enhances production efficiency, reducing product costs and stimulating consumer demand. This reshapes the demand side and further drives the development of related industries. Industrial structural upgrading, characterized by its strong innovative attributes, facilitates the transition from basic manual labor to complex intelligent labor. It promotes inter-industry collaborative innovation, advances leaps in production technologies, creates new growth opportunities, and fosters competitive advantages, thereby driving the development of novel productive forces. Consequently, we propose Hypothesis 2c:

H2c: Digital infrastructure construction promotes the development of novel productive forces by enabling industrial structural upgrading.

Spatial spillover effects of digital infrastructure construction on new quality productivity

The improvement of digital infrastructure can break temporal and spatial constraints, facilitating closer connections between regions and enhancing communication and cooperation among regions, thereby strengthening the dynamic interactions that drive the development of new quality productivity. In other words, the results in spatial spillover affect the development of new quality productivity in neighboring areas, thereby promoting such productivity growth in adjacent regions. Moreover, enhanced digital infrastructure facilitates the cross-regional flow of resources, enabling efficient allocation and movement of resources and reduces the spatial transportation costs associated with inter-regional resource flow40. By leveraging demonstration and sharing effects, improved infrastructure can further stimulate the development of new quality productivity in surrounding areas. Therefore, this paper posits the following hypothesis:

H3: The construction of digital infrastructure can exert positive spatial spillover effects, promoting the development of new quality productivity in neighboring cities.

Data and empirical methodology

Model design

To examine the impact of digital infrastructure construction on new quality productivity, we construct the following econometric model:

$$NQPi,t=\beta 0+\beta 1 \times infi,t+\theta 1 \times Xi,t+\varepsilon i,t$$
(1)

where \(NQP\) represents new quality productivity, \(inf\) denotes digital infrastructure, X includes control variables, and \(\varepsilon\) represents the random disturbance term, with subscripts i and t indicating city and year, respectively.

Based on the theoretical analysis in the preceding sections, this study posits that the development of digital infrastructure promotes the advancement of new productive forces through three key pathways: enhancing technological innovation capabilities, improving the efficiency of resource allocation, and facilitating industrial structure upgrading. First, Eq. (2) is formulated to examine whether digital infrastructure development impacts the mediating variables. Second, to address potential concerns regarding the adequacy of theoretical justification for the mediating variables’ effects on new productive forces, Eq. (3) is developed to test the influence of the mediating variables on new productive forces. To empirically test these pathways, the study constructs the following models:

$$Mi,t=\alpha 0+\alpha 1 \times infi,t+\theta 2 \times Xi,t+\varepsilon i,t$$
(2)
$$NQPi,t=\gamma 0+\gamma 1 \times Mi,t+\theta 3 \times Xi,t+\varepsilon i,t$$
(3)

Where M represents the mediating variables, which in this study include technological innovation capabilities, resource allocation efficiency, and industrial structure upgrading.

Variable descriptions

Dependent variable

New quality productivity (NQP)

We comprehensively measure NQP from two dimensions: substantive elements and pervasive elements.

Substantive elements encompass new quality labor force, new quality labor resource, and new quality labor object, representing tangible and perceivable forms of existence in material creation. A new quality labor force refers to a workforce that has undergone skill training, possesses modern scientific and technological knowledge, and is proficient in utilizing advanced tools. It serves as the most dynamic component of new productivity. This study represents the latest labor force through three dimensions: the number of employees in emerging industries, individual employee capabilities, and the proportion of high-quality employees. New quality labor resources include the new tools directly involved in the production process and the novel infrastructure indirectly supporting it. These factors act as significant drivers for the development of new productivity. The study evaluates new labor means based on infrastructure, future development potential, and the ecological environment. New quality labor objects form the foundation for advancing new productivity. With continuous advancements in science and technology, the scope of new labor objects has expanded beyond traditional ones. Five dimensions reflect new labor objects: scientific research and development, innovative outputs, intelligence, greening, and data-driven elements.

Pervasive elements include technological innovation, factor innovation, and organizational innovation, which primarily manifest as virtual material forms centered around new technologies. Technological innovation highlights the enthusiasm of prefecture-level cities in conducting innovative activities. The study uses the number of patents granted in these cities as a proxy for technological innovation. Factor innovation emphasizes the investment efforts of prefecture-level cities in education, science and technology. The indicators employed include expenditures on science and technology and education. Organizational innovation underscores the development of the tertiary sector. The study measures organizational innovation through the growth of the service industry.

Based on these two dimensions, this study constructs a comprehensive evaluation system consisting of six primary and fifteen secondary indicators, as shown in (Table 1). The entropy method is applied to quantify the indicators for new productivity.

Table 1 Measurement indicator system for new quality productivity.
Full size table

Core explanatory variable

Digital infrastructure (DI)

Given that digital infrastructure is a multi-layered composite indicator, a single metric cannot adequately reflect the development level of digital infrastructure across various cities. We select six secondary indicators from both input and output perspectives, employing the entropy method to measure digital infrastructure construction in different cities. The indicator measurement system is presented in (Table 2).

Table 2 Measurement indicator system for digital infrastructure.
Full size table

Mediating variable

Technological innovation capability (TIC)

This study uses various prefecture-level cities’ innovation and entrepreneurship index as a proxy for a city’s technological innovation capability.

Factor allocation efficiency (FAE)

Each prefecture-level city’s factor market distortion index is utilized to measure factor allocation efficiency. First, the marginal output of capital and labor is divided by their respective prices to compute the factor market distortion index. A distortion index greater than 1 indicates lower factor allocation efficiency, while an index less than or equal to 1 indicates higher efficiency. Second, to compare regression results in mechanism testing, the reciprocal of the distortion index is taken if the value is greater than 1. No transformation is applied if the distortion index falls between 0 and 1. This adjusted value serves as an alternative indicator of factor allocation efficiency.

Industrial structure upgrading (ISU)

This study uses the Theil index to measure the rationalization of city industrial structures. The Theil index serves as a proxy for industrial structure upgrading, and its calculation formula is as follows:

$$ISUi,t=\sum\nolimits_{{m=1}}^{3} {yi,m,t\ln \left( {{{yi,m,t} \mathord{\left/ {\vphantom {{yi,m,t} {li,m,t}}} \right. \kern-0pt} {li,m,t}}} \right)}$$
(4)

where \(yi,m,t\) represents the added value of the mth industry in city i during period t, while \(li,m,t\) denotes the proportion of employees in the mth industry to the total employment in city i during the same period.

Control variables

We also select other control variables, which primarily include:

  • Financial development level (FDL). The ratio of the year-end loan balance of financial institutions in each prefecture-level city to its GDP is used as a proxy for financial development.

  • Infrastructure development (ID). Per capita urban road area serves as a proxy for the level of infrastructure.

  • Government intervention (GI). The ratio of government fiscal expenditure to GDP is employed as a measure of government intervention.

Data sources

This study focuses on 276 prefecture-level cities in China from 2011 to 2020 to empirically examine the impact and channels through which digital infrastructure affects new quality productivity. Most of the research data were sourced from the China City Statistical Yearbook, various city statistical yearbooks, the National Intellectual Property Administration website, Tianyancha, government work reports, and manual data collection. Missing values were addressed using linear interpolation. Descriptive statistics for the main variables are presented in (Table 3).

Table 3 Descriptive statistics for variables.
Full size table

Empirical results and analysis

Empirical results of the baseline regression

This study employs a stepwise approach to incorporating control variables into Model (1) to examine the development of new productive forces empowered by the construction of digital infrastructure, as illustrated in (Table 4). The regression results indicate that, regardless of the inclusion of control variables, the coefficient for the core explanatory variable, digital infrastructure, is consistently significant and positive. Focusing on column (5), the coefficient for digital infrastructure is 0.055, which implies that improvements in digital infrastructure lead to an average increase of 0.055% in new quality productivity. The finding supports the hypothesis that the development of digital infrastructure has significant economic implications for enhancing new quality productivity.

Based on the regression results of the control variables, the regression coefficients of financial development level are significantly positive, indicating that improvements in financial development can provide crucial funding support and empower the growth of new quality productivity. In contrast, the regression coefficients of infrastructure development are significantly negative, suggesting that the improvement of certain infrastructure may depend on high-pollution and high-energy-consuming industries, which hinders the development of new quality productivity. Additionally, the regression coefficients of government intervention are significantly positive, demonstrating that government intervention supports the growth of new quality productivity.

Table 4 Empirical results of baseline regression.
Full size table

Empirical results of the robustness analysis

To ensure the robustness of the baseline regression results, this study employs several methods: excluding the impact of the pandemic, omitting municipalities directly under the central government, and utilizing a difference-in-differences approach for robustness testing. The specific methods and regression outcomes are as follows:

Excluding the impact of the pandemic

Given the significant disruption caused by the COVID-19 pandemic on China’s economic development in 2020, which could potentially affect the research results, we excluded data from that year and re-conducted the regression analysis to ensure the robustness of the baseline results. As shown in column (1) of (Table 5), the regression results indicate that the coefficient for digital infrastructure remains significantly positive. This suggests that even when excluding the impact of the pandemic, the improvement of digital infrastructure still contributes to the development of new productive forces, confirming the robustness of the baseline results.

Omitting municipalities directly under the central government

Considering the unique political, economic, and cultural significance of China’s four municipalities directly under the central government, which exhibit markedly superior economic development levels, technological innovation capabilities, and access to preferential policies compared to other prefecture-level cities, we omitted data from these municipalities to ensure the robustness of the baseline regression results. As shown in column (2) of (Table 5), the regression results reveal that the coefficient for digital infrastructure remains significantly positive, indicating that even after excluding the influence of these municipalities, improvements in digital infrastructure continue to promote the development of new productive forces, further confirming the robustness of the baseline results.

Difference-in-differences approach

Recognizing that numerous scholars have utilized the “Broadband China” strategy to assess the policy effects of digital infrastructure development, this study employs a multi-period difference-in-differences model for robustness testing. Specifically, we first identified the pilot cities for the “Broadband China” initiative, determined in batches in 2014, 2015, and 2016, as the treatment group (treat = 1), while non-pilot cities served as the control group (treat = 0). Next, the years in which the “Broadband China” pilot policy was implemented and subsequent years were designated as time = 1, while the years before the policy implementation and the non-pilot cities were set as time = 0. Finally, we constructed a multi-period difference-in-differences model by incorporating the interaction term of treat and time as a substitute variable for digital infrastructure and re-conducted the regression analysis, as shown in column (3) of (Table 5). The regression results demonstrate that the coefficient for digital infrastructure remains significantly positive, suggesting that even when applying the difference-in-differences method, improvements in digital infrastructure still contribute to the development of new productive forces, affirming the robustness of the baseline results.

Considering the impact of other policies

During the sample study period, other policies, such as the pilot program for innovative cities (PPIC), talent attraction policies (TAP), and the national pilot zone for innovative digital economy development (NPZIDEP), may also have influenced the development of new productivity in prefecture-level cities. Thus, this study incorporates dummy variables for the policies mentioned above into the baseline model for regression analysis. The dummy variables are defined as follows: before the implementation of a policy, the variable is set to 0, while in the year of implementation and subsequent years, it is set to 1. The regression results are shown in column (4) of (Table 5). The results indicate that the regression coefficient of digital infrastructure remains significantly positive. The finding suggests that even after accounting for the influence of these additional policies on new productivity development, the improvement of digital infrastructure continues to promote new productivity to a certain extent, thereby confirming the robustness of the baseline regression results.

Table 5 Empirical results of the robustness.
Full size table

Empirical results of the endogeneity check

Although the benchmark regression model has fixed individual and time effects, omitted variable bias may still lead to endogeneity issues within the model. We employ the instrumental variable method to mitigate potential endogeneity concerns. The number of post offices in Chinese prefecture-level cities in 1984 is chosen as an instrumental variable for digital infrastructure. The reasons for selecting this instrument are twofold: First, it meets the relevance requirement. The development of digital infrastructure relies on modern information technologies such as 5G and big data. The number of post offices in 1984 reflects information technology development in various prefecture-level cities during the late 20th century, which directly influences the advancement of contemporary information technologies, thus satisfying the relevance criterion. Second, it fulfills the homogeneity requirement. The number of post offices in 1984 is historical data unlikely to affect the development of new productive forces, thereby satisfying the homogeneity criterion.

Moreover, considering the number of post offices in 1984 consists of cross-sectional data, estimating in a fixed effects model is challenging. Therefore, this study employs the interaction term between the number of post offices in 1984 and a time dummy variable as an instrumental variable for the endogeneity test. The results of the instrumental variable test, presented in Table 6, indicate that both the identification and weak instrument tests reject the null hypothesis, confirming the absence of issues related to identification and weak instruments.

The results of the endogeneity test using two-stage least squares regression are shown in (Table 7). These results demonstrate that the regression coefficient for digital infrastructure remains significantly positive, indicating the robustness of the benchmark model’s findings.

Table 6 Results of instrumental variable tests.
Full size table
Table 7 Empirical results of endogeneity check.
Full size table

Empirical results of the heterogeneity test

The baseline regression results indicate that the development of digital infrastructure significantly contributes to advancing emerging productivity. Considering that the magnitude of this effect may vary depending on city size, administrative level, and regional distribution, this study conducts heterogeneity analyses from these three perspectives.

Heterogeneity by city size

To examine the impact of city size, the study calculates the average population of sampled cities between 2011 and 2020 and classifies cities into two groups based on the median population. Cities with populations above the median are categorized as larger cities, while those below the median are categorized as smaller cities. Separate regression analyses are conducted for each group, and the results are presented in (Table 8). The findings reveal that in smaller cities, the regression coefficient of digital infrastructure is significantly positive, indicating that digital infrastructure development promotes emerging productivity in these areas. However, in larger cities, the coefficient is not statistically significant, suggesting a negligible impact on emerging productivity. The underlying reason may lie in the higher degree of factor agglomeration in larger cities, which can lead to congestion effects that diminish the benefits of digital infrastructure for productivity enhancement.

Heterogeneity by administrative level

For the analysis of administrative level, cities are divided into two groups: key cities (including municipalities directly under the central government, provincial capitals, and sub-provincial cities) and non-key cities (all other cities). Separate regression analyses are conducted, and the results are shown in (Table 8). The results indicate that in key cities, the regression coefficient of digital infrastructure is significantly positive, suggesting that digital infrastructure development fosters emerging productivity. Conversely, in non-key cities, the coefficient is significantly negative, implying that digital infrastructure development may hinder productivity growth. This could be attributed to the relatively lower economic development levels and inadequate institutional environments in non-key cities, which may lead to unhealthy competition and undermine the positive impact of digital infrastructure.

Heterogeneity by regional distribution

To analyze regional differences, cities are classified into eastern and central-western regions based on their geographic location. Regression analyses are performed separately for these two groups, with the results summarized in (Table 8). The findings reveal that in eastern cities, the regression coefficient of digital infrastructure is not statistically significant, indicating a limited impact on emerging productivity. In contrast, in central-western cities, the coefficient is significantly positive, suggesting that digital infrastructure development drives productivity growth in these regions. This disparity can be explained by the more advanced digital infrastructure in eastern cities, where the marginal benefits have diminished over time. Conversely, in central-western cities, digital infrastructure development is still in its adoption and imitation phase, benefiting from latecomer advantages and higher marginal returns, which enable it to contribute significantly to productivity advancement.

Table 8 Empirical results of heterogeneity check.
Full size table

Empirical results of the mechanism check

The regression analysis above confirmed the positive role of digital infrastructure development in enhancing the new productivity quality. To further validate the mechanisms of this influence, we conducted regression tests based on Eqs. (2) and (3). The results are presented in (Table 9).

Mechanism of technological innovation capability

Columns (1) and (2) in Table 9 display the regression results for the mechanism of technological innovation capability. In Column (1), the regression coefficient of digital infrastructure is significantly positive, indicating that the development of digital infrastructure contributes to enhancing urban technological innovation capability. In Column (2), the regression coefficient of technological innovation capability is also significantly positive, suggesting that improvements in technological innovation capability promote the development of new productivity quality. Therefore, the results in Columns (1) and (2) demonstrate that digital infrastructure enhances new productivity quality by fostering urban technological innovation capability.

Mechanism of factor allocation efficiency

Columns (3) and (4) in Table 9 illustrate the regression results for the mechanism of factor allocation efficiency. In Column (3), the regression coefficient of digital infrastructure is significantly positive, showing that digital infrastructure development improves factor allocation efficiency. Similarly, in Column (4), the regression coefficient of factor allocation efficiency is significantly positive, indicating that improved factor allocation efficiency facilitates the development of new productivity quality. These results suggest that digital infrastructure empowers new productivity quality by enhancing factor allocation efficiency, as shown in Columns (3) and (4).

Mechanism of industrial structure upgrading

Columns (5) and (6) in Table 9 present the regression results for the mechanism of industrial structure upgrading. In Column (5), the regression coefficient of digital infrastructure is significantly positive, confirming that digital infrastructure development drives industrial structure upgrading. In Column (6), the regression coefficient of industrial structure upgrading is significantly positive, demonstrating that the advancement of industrial structure supports the development of new productivity quality. Thus, the findings in Columns (5) and (6) indicate that digital infrastructure promotes new productivity quality through industrial structure upgrading.

Table 9 Empirical results of mechanism check.
Full size table

Further analysis: spatial spillover effects of digital infrastructure development

The regression analysis above verifies the promoting effect of digital infrastructure development on new quality productivity and its mechanisms of action. Considering the externalities of digital infrastructure development and the spatial dependency in developing new quality productivity among cities, this study further draws on the approach of Sunak et al.41 to construct a spatial durbin model (5). The model is employed to examine the spatial spillover effects of digital infrastructure development on the new quality productivity of neighboring cities.

$$NQPi,t=\rho \sum WijNQPi,t+\beta \inf i,t+\gamma \sum Wij\inf i,t+\theta Xi,t+\varepsilon i,t$$
(5)

where \(\rho\) represents the spatial lag coefficient of new quality productivity, \(\beta\) denotes the impact of digital infrastructure development on local new quality productivity, and \(\gamma\) signifies the effect of digital infrastructure development on the new quality productivity of nearby areas.

$$W_{{i,j}}^{a}=\left\{ \begin{gathered} 0,i=j \hfill \\ \frac{1}{{d_{{i,j}}^{2}}},i \ne j \hfill \\ \end{gathered} \right.$$
(6)

where \(W_{{i,j}}^{a}\) represents the geographic distance matrix, \(d_{{i,j}}^{{}}\) indicates the geographic distance between city i and city j.

$$W_{{i,j}}^{b}=\left\{ \begin{gathered} 0,i=j \hfill \\ \frac{1}{{\left| {GD{P_i} - GD{P_j}} \right|}},i \ne j \hfill \\ \end{gathered} \right.$$
(7)

where \(W_{{i,j}}^{b}\) denotes the economic distance matrix, \(GD{P_i}\) and \(GD{P_j}\) represent the GDP of city i and city j, respectively.

Table 10 presents the regression results of the spatial Durbin model, using the geographic distance matrix as an example. The significant positive spatial lag coefficient (ρ) confirms a notable positive spatial autocorrelation of new productivity among different cities. Furthermore, the indirect effect shows that the regression coefficient for digital infrastructure development is significantly positive, indicating that the advancement of digital infrastructure promotes the development of new productivity locally and generates a positive spatial spillover effect on the latest productivity of surrounding areas. Finally, the robustness of the conclusion is reaffirmed through the use of an economic distance matrix.

Table 10 Suitability test for the spatial durbin model.
Full size table

Research conclusions and policy implications

The construction of digital infrastructure and the development of new quality productivity are intrinsic requirements and critical focus areas for driving innovation in the digital economy and achieving high-quality economic growth. Effectively integrating digital infrastructure construction with the advancement of new quality productivity has profound implications for reshaping China’s economic landscape. This study examines the impact of digital infrastructure construction on new quality productivity using empirical analysis based on data from 276 prefecture-level cities in China from 2011 to 2020. The key findings are as follows: First, digital infrastructure construction significantly promotes the development of new quality productivity. Second, the effects of digital infrastructure construction on new quality productivity vary across cities of different sizes, administrative levels, and geographical regions. In smaller cities, digital infrastructure construction positively influences new quality productivity, whereas in larger cities, it has no significant impact. For cities of different administrative levels, digital infrastructure construction fosters new quality productivity in key cities but hinders its development in non-key cities. Geographically, digital infrastructure construction has no significant impact in eastern cities but promotes new quality productivity in central and western cities. Third, digital infrastructure construction enhances new quality productivity primarily through mechanisms such as boosting technological innovation, improving resource allocation efficiency, and driving industrial structural upgrades. Lastly, digital infrastructure construction not only promotes local new quality productivity but also generates positive spatial spillover effects, benefiting neighboring regions.

The findings of this study offer significant policy implications for accelerating the development of China’s digital economy to drive high-quality growth: (1) Accelerating the construction of digital infrastructure and exploring new pathways for advanced productivity development. The government should guide and promote digital infrastructure construction in an orderly manner, prioritizing the development of digital infrastructure in central cities to maximize their radiative impact on surrounding regions. Efforts should focus on integrating digital technologies into residents’ lives, corporate production, and government oversight to foster the growth of advanced productivity. (2) Adopting a differentiated approach to digital infrastructure construction based on urban characteristics. Recognizing that the impact of digital infrastructure on advanced productivity varies by city size, level, and regional distribution, policies should avoid a “one-size-fits-all” approach. Instead, construction should align with local economic development levels and industrial needs, minimizing resource underutilization. For digital infrastructure linked to smart environments, logistics, and living, efforts should accelerate the implementation of these systems to enhance integration and establish sustainable, clean service models. For technologies like 5G, big data, and artificial intelligence, investment should be increased, and cooperation among network operators should be encouraged to capitalize on digital dividends. (3) Enhancing technological innovation, resource allocation efficiency, and industrial upgrading as drivers of advanced productivity. The government should strengthen the protection of scientific and technological achievements through digital mechanisms, improving intellectual property protection, management, and oversight. It would foster a healthy market environment, motivating enterprises to innovate and drive productivity growth. Additionally, promoting the integration of new and traditional factors of production and fostering cross-enterprise and cross-industry collaboration can enhance market efficiency. Establishing a well-functioning factor market with reduced transaction costs will further support productivity. Finally, accelerating the integration of digital technologies into traditional industries and the service sector can unlock digital dividends and empower advanced productivity development. (4) Strengthening the radiative and catalytic role of digital infrastructure in neighboring regions. By increasing network density and node numbers, the government can break down barriers to resource flow between regions, promote interaction between central cities and surrounding areas, and expand the radius of knowledge and technology spillovers, thus fostering overall regional productivity growth.