Spatiotemporal evolution characteristics and influencing factors of digital industry in China Authors: Hong Zhang, Fengjiao Li, Suqiong Wei, Lili Jiang, Junfeng Xiong, Tianwei Zhang Published: 2024 ABSTRACT AbstractIn the era of the global knowledge economy, the rapid advancement of new-generation information technologies has positioned digital industry at the forefront of socio-economic development. Digital industrial clusters, comprising key enterprises and related institutions, have emerged as global hubs for technological innovation, drawing recognition from governments and international organizations. However, China’s digital industry faces challenges characterized by being “large but not strong” and experiencing “industrial structure softening.” This paper explores the temporal evolution and spatial distribution of China’s digital industry, shedding light on its unique patterns. This study reveals that China’s digital industry has progressed through four distinct phases, evolving from manufacturing-oriented to service-oriented, and experiencing dynamic stability. In terms of spatial distribution, it exhibits a pattern known as the “One Core, Three Poles, and Multiple Strongholds,” with Beijing as the core and Shanghai, Shenzhen, Hangzhou, among others, serving as growth poles. The study identifies factors influencing spatial distribution, including software development, patent authorizations, mobile phone penetration, industrial parks, digital finance, and foreign investment. Manufacturing-based digital industries prioritize innovation and IP protection, while service-based sectors rely on software development, innovation, and international collaboration. The GWR model demonstrates spatial variation, emphasizing the importance of tailored development strategies. This research contributes to understanding the dynamics of China’s digital industry and provides valuable insights for policymakers aiming to create an internationally competitive digital industry cluster.Keywords: Digital industry, Listed enterprises, Geographical weighted regression, ChinaSubject terms: Environmental social sciences, Sustainability Introduction As information and communication technologies, along with data processing technologies, continually evolve, the rapid advancement of the new generation of information technologies—exemplified by the Internet of Things (IoT), cloud computing, big data, artificial intelligence (AI), augmented reality (AR), and virtual reality (VR)—is elevating the digital industry to a more prominent role in the integration of digital technology with socio-economic development1. Digital industrial clusters, recognized as global hubs for technological innovation, receive widespread acclaim from governments and international organizations2. These clusters consist of key enterprises in the digital industry and their associated institutions, making them prominent players in the realm of digital technology3. Notable examples include Silicon Valley in the USA, which focuses on microelectronics, Japan’s Tsukuba Science City, known for multidisciplinary integration, South Korea’s Daejeon Techno Valley, representing the consumer electronics sector, and India’s Bangalore Information Technology Center, famous for its software industry cluster4. With the advent of the global knowledge economy era, the development of the digital industry is placing increasing emphasis on the significance of digital technology. It is also emerging as a critical force in the global realignment of resources, the restructuring of the global economic framework, and the transformation of the competitive landscape. Compared to traditional industries, the digital industry differs significantly from traditional industrial clusters in terms of geographic resource requirements, clustering drivers, and development pathways. From a literature review perspective, research on high-tech industries is relatively well-established. However, studies on digital industries within the broader context of the digital economy remain limited. First, digital industry clusters are more prominent in developed cities, and existing research primarily focuses on “first-tier” regions globally, such as Silicon Valley, the Yangtze River Delta, and Singapore, with limited analysis of digital industry development across China. Second, most studies concentrate on specific types of industries or high-tech intelligent sectors, with few examining the spatial patterns of digital industries from a digital industrialization perspective. Finally, there is relatively little research addressing the heterogeneity among different types of digital industries. For China, there are currently challenges within its digital industry, characterized by being “large but not strong” and experiencing “industrial structure softening.” Several key issues require attention: What has been the evolution of China’s digital industry so far? What are the trends in its overall spatial layout, and what are the characteristics of digital industry development at different stages? Are there notable differences among various types of digital industries? Additionally, it is crucial to examine the factors influencing the spatial distribution of China’s digital industry, how to optimize its national layout for greater efficiency and balance, and how to build “internationally competitive digital industry clusters” that foster collaborative, open, and innovative development. In-depth research on these issues will offer valuable insights for policy implementation and industrial growth while providing a uniquely Chinese perspective on exploring the spatial patterns of the digital industry. Based on a review and synthesis of existing research, this study utilizes data from core enterprises in the digital economy and focuses on China’s digital industry. It analyzes the spatiotemporal patterns of China’s digital industry, identifies the dynamic evolutionary patterns of different types of digital industries, and explores the factors influencing their spatial distribution, ultimately providing scientifically grounded policy recommendations for the development of China’s digital industry. The paper is organized as follows. A comprehensive literature review of digital industry is presented in the second section. The third section describes the methodology. The detailed account of the results obtained is described in the fourth section. The fifth section encompasses the conclusions drawn from the study and offers a discussion on the findings. Literature review The digital industry concept: core industries of the digital economy Significant technological breakthroughs in microelectronics, marked by the invention of the second-generation transistor electronic computer and integrated circuits, paved the way for early explorations of the digital economy. In 1962, Fritz Machlup highlighted the importance of businesses providing information products or services and introduced the concept of the “information economy.” With the rapid spread of digital technologies, the concept of the information economy gained wide recognition and its scope continuously expanded5. In 1994, renowned new economist Don Tapscott firstintroduced the term “digital economy”6. As information and communication technology and the digital economy continued to develop, international organizations and several national statistical agencies conducted a series of studies on the content and classification of the digital economy7. Currently, research on the digital industry primarily focuses on three aspects: the development of the digital industry, its impact on regional economies, and the integration of industry and digitization. In terms of development, Tatarinov8 explored methods and practical issues in evaluating the digital economy within the macroeconomic context, suggesting that digital economy accounting should be considered from both taxation and welfare perspectives. Regarding its impact on regional economies, Degryse9 argued that the “digital” era has transformed production and daily life, and the digital economy, as a new engine of economic growth, brings disruptive changes to related industries. Regarding the integration of industry and digitization, Barker10 focused on the application of technology in the telecommunications sector, highlighting the New Zealand government’s efforts to explore digital integration. Industrial evolution In the realm of industrial cluster research, during its early stages, the attention of industrial cluster research was primarily given to issues related to cluster formation, intrinsic mechanisms, and industrial location choices. Research on industrial clusters primarily centered around the mechanisms of industrial clusters, technological innovation, organizational innovation, social capital, and the relationship between economic growth and industrial clusters. During the late 19th century and early 20th century, the founder of the neoclassical school of economics, Marshall, put forth the argument that industrial clustering results in external economic advantages11. In 1934, economist Joseph Schumpeter introduced the concept of “creative destruction,” where innovation drives economic growth through the continuous introduction of new products, technologies, and business models. Later studies refined Schumpeter’s ideas, distinguishing two innovation patterns: Schumpeter Mark I, led by new entrants, and Schumpeter Mark II, dominated by established firms12. These patterns have been observed in various sectors, including ICT, where high-growth applications exhibit high rates of innovation and diversified knowledge bases, while low-growth applications show less diversification and lower entry rates13. Additionally, co-evolutionary patterns in sectors like mining highlight how knowledge networks and global value chains evolve in divergent ways, with knowledge networks remaining stable while trade networks become more fragmented14. These studies underscore how technological change shapes industrial dynamics and productivity growth15,16. In 1979, Italian economist Becattini revealed characteristics of the “new industrial district,” including geographical proximity, sector specialization, dominance of small and medium-sized enterprises, collaborative enterprise division of labor, shared social culture, and regional rootedness17. Subsequently, the evolution of clusters and regional development emerged as significant research hotspots, influenced by the “relational turn” and “evolutionary turn” in Western economic geography. Scholars began to investigate the mechanisms of clustering evolution and its influencing factors, emphasizing concepts such as rootedness, path dependency, and multidimensional proximity. In 1999, Ahokangas proposed an evolutionary model of regional industrial clusters, dividing the cluster life cycle into three stages: origin and emergence, growth and concentration, and maturity and regeneration18. In 2010, Menzel regarded enterprise knowledge heterogeneity as the core driving force of cluster evolution19. Currently, key research topics concerning industrial clusters focus on the interplay between technology, knowledge, innovation, and geographical space. Researchers examine the effects of cluster networks and innovation through the lenses of local buzz and global pipelines, collective learning, and knowledge spillovers20. Research on digital industrial clusters Digital industry clusters, as a burgeoning model of industrial organization, have a primary objective of fostering regional economic growth. These clusters strategically bring together digital enterprises in close proximity, enabling them to harness collective advantages such as shared suppliers, the cultivation of skilled workforces, the stimulation of robust competitive dynamics, and, as a result, the enhancement of local competitiveness within the digital industry21. From a methodological perspective, research on digital industry clusters by scholars primarily focuses on their development, often employing methods such as expert scoring and semi-structured interviews. For instance, Larissa Tashenova22 assesses the digital potential of core innovative industrial clusters through expert scoring. Similarly, Max Nathan and his colleagues23 employ semi-structured interviews to investigate London’s digital economy and technology city. However, existing studies predominantly use qualitative methods and concentrate on small-scale areas, with a notable lack of quantitative research that addresses broader regional perspectives. From the perspective of research content, scholars have focused their studies on specific subcategories within the digital industry, such as the agglomeration effects of the Information and Communications Technology (ICT) manufacturing sector and the internet services sector. Research on specific industries within this field generally falls into two categories: One is the manufacturing-oriented digital industries, such as Information and Communications Technology (ICT). For instance, Østergaard et al.24 studied the decline of Denmark’s wireless communications cluster, while Harbi et al.25 examined the development of high-tech zones and the necessary conditions for growth, using Tunisia’s ICT industry as a case study. Regarding manufacturing-oriented digital industries, Chinese scholars Song26 explored the spatial hierarchy of newly established communication equipment manufacturing enterprises in China. Overall, the spatial evolution of high-end manufacturing industries, such as electronic information, shows heterogeneity, with a trend toward suburbanization in their agglomeration patterns27. The other is the service-oriented digital industries, such as internet and information technology services. For instance, the research by Guo28 found that the development level of e-commerce in prefecture-level cities in China generally exhibits a hierarchical diffusion characteristic of “large agglomeration with small dispersion.” Ernkvist29 and his colleagues have studied the decline of the Japanese game software industry, while Mathijs de Vaan30 has conducted research on the global video game industry. This category is characterized by the “intermediate goods” nature of the industry, which results in a pattern of “large agglomeration and small dispersion.” Additionally, there is a tendency for the industry’s structure to evolve from a single-core to a multi-center configuration31,32. This study contributes to the existing literature by addressing the gap in research on digital industry through a more nuanced methodological approach and a broader content focus. Previous research has often relied on qualitative methods such as expert scoring and semi-structured interviews to examine the development of digital industry clusters. This study introduces a quantitative perspective by integrating advanced spatial analysis methods and Geographical Information Systems (GIS). Additionally, whereas existing studies have predominantly concentrated on specific subcategories, such as ICT manufacturing and internet services, this research provides a comprehensive analysis of the entire digital industry, encompassing both manufacturing and service-oriented digital industries. By expanding the methodological toolkit and research scope, this study provides valuable guidance for policy implementation and industrial development, offering a uniquely Chinese perspective on the spatial patterns and dynamics of the digital industry. Methodology Data collection and procedures The data processing operations were conducted as follows: First, we established a data pool of all listed digital industry enterprises from Tianyancha, capturing information from 1980 to 2020. We utilized the “Statistical Classification of the Digital Economy and its Core Industries (2021 Edition)” to define “digital industrialization.” This classification includes 20 secondary industry categories related to the digital sector as outlined in the National Economic Industry Classification and Codes (GB/T 4754 − 2017), which guided our identification of relevant companies. To retrieve companies within these classifications, we accessed the reliable third-party enterprise information platform, Tianyancha (URL: https://www.tianyancha.com/), which sources data from government records and company disclosures, ensuring extensive coverage across various industries and regions33. Our selected listing statuses encompass A-shares, H-shares, Hong Kong-listed stocks, the Sci-Tech Innovation Board, the New Third Board, and the New Fourth Board. Compared to typical digital enterprises, listed firms tend to exhibit greater economic scale, brand value, and market influence, positioning them as leaders in China’s digital sector34. This initial search yielded 10,238 entries, excluding enterprises from Hong Kong, Macao, and Taiwan. We collected comprehensive data, including company names, business status, establishment dates, industry fields, registered addresses, and business scopes. Based on our defined digital industry criteria, we filtered out non-relevant companies, ultimately organizing the remaining 8,566 enterprises by industry and administrative region—5,561 service-oriented and 2,975 manufacturing enterprises. Figure 1 illustrates their geographical distribution. Next, we used the Baidu Map API to obtain geographic coordinates in the WGS-84 system, prioritizing addresses from the latest annual reports and supplementing with registered addresses where necessary. Finally, we analyzed this data to uncover spatial patterns and mechanisms within China. Spatial distribution map of China’s digital industry listed companies. Additionally, we gathered relevant socioeconomic data and digital economy indices. Socioeconomic data on business revenue, employment figures, and digital product quantities per province were sourced from the “China Statistical Yearbook” and various provincial yearbooks (URL: http://www.stats.gov.cn/). National development zone data came from the “Catalog of Chinese Development Zone Review Announcements” (2018 Edition) (URL: https://www.gov.cn/xinwen/2018-03/03/content_5270330.htm). Digital economy indices, such as the Digital Innovation Patent Authorization Index, registered software copyrights, and foreign investment attraction scores, were obtained from the Enterprise Big Data Research Center at Peking University (URL: https://cer.gsm.pku.edu.cn/). The Digital Inclusive Finance Index was sourced from the Digital Finance Research Center at Peking University (URL: https://idf.pku.edu.cn/). Annual statistical data for digital industry sectors were derived from the Ministry of Industry and Information Technology’s Operations Monitoring and Coordination Bureau (URL: https://www.miit.gov.cn/jgsj/yxj/). Spatial pattern and evolution This study utilized a range of spatial analysis methods to conduct an in-depth examination of the development characteristics of China’s digital industry. Specifically, the Nearest Neighbor Index (NNI) was used to quantitatively assess the overall spatial distribution of the industry, while the Standard Deviational Ellipse method analyzed the trajectory and primary directional trends of the industry’s spatial development. Following this, Kernel Density Estimation (KDE) was used to identify and analyze the spatial distribution patterns, and Spatial Autocorrelation analysis was used to evaluate the degree of global spatial autocorrelation. Compared to other methods, these choices offer a deeper understanding of the complexity of industry evolution and regional differences. A brief overview of the formulas and principles underlying each method is provided below: The Nearest Neighbor Index (NNI) is employed to assess the spatial distribution pattern of point features. It works by calculating the average distance between each spatial point feature of digital industry enterprises and its nearest neighbor spatial point feature. In this analysis, the spatial coordinates of enterprises collected during the third step of the previous process represent the spatial point features of digital industry enterprises. This measurement is then compared to the theoretical average distance of digital industry enterprise spatial point features, assuming a random distribution. This process enables the evaluation of the spatial distribution pattern of point features within digital industry enterprises35. The formula for computing the NNI is as follows: NNI represents the Nearest Neighbor Index. represents the observed average distance between spatial point features of digital industry enterprises and their nearest neighbor spatial point features. is the expected average distance between specified spatial point features of digital industry enterprises under a random pattern. di denotes the actual distance between enterprise “i” and other enterprises. n is the total number of digital industry enterprises, and A stands for the area of the study region. If NNI is less than 1, it indicates an agglomerated distribution of digital industry enterprises in the city. If NNI is greater than 1, it suggests a dispersed distribution of digital industry enterprises in the city. If NNI equals 1, it signifies a random spatial distribution of digital industry enterprises in the city. The Standard Deviational Ellipse method, which describes characteristics such as centrality, dispersion, orientation, and shape of an ellipse, can be employed to unveil the overall features and dynamic evolution trends in the spatial distribution of China’s digital industry36. The key parameters include centroid coordinates, area, major axis lengths (xstdDist) and minor axis lengths (ystdDist), and orientation angle. Among these parameters, centroid coordinates indicate the relative positions of China’s digital industry enterprise spatial distribution, while the area signifies the extent of dispersion in the digital industry spatial distribution. The xstdDist and ystdDist respectively represent the degree of dispersion of digital industry enterprises in the primary and secondary directions, and the orientation angle reflects the primary trend direction of China’s digital industry distribution. The Kernel Density Estimation (KDE) is primarily a method for estimating probability density values based on the distances between points to be estimated and sample points within a given region. It leverages the spatial attributes of the data samples to study spatial distribution characteristics and effectively reveals the intensity of clustering in the spatial distribution of digital industry enterprises36. It is suitable for visually representing the distribution patterns of point features in space. In this article, the KDE method is used, treating digital industry enterprises as “points” in space, to analyze the spatial distribution patterns and clustering variations of China’s digital industry enterprises37. The expression for the Kernel Density Analysis function is as follows: Where k(x) represents the default spatial weight kernel function, h (h > 0) is the analysis range threshold38, and (x-xi) represents the distance from the estimation point x to the observation xi. Spatial autocorrelation analysis is used to measure to what extent a geographic unit’s attributes depend on neighboring units’ attributes39. Global and local autocorrelation indices assess the level of clustering or dispersion in spatial data. Global spatial autocorrelation indices mainly determine if there is clustering or dispersion among the spatial elements of China’s core digital economy-listed enterprises. A higher Global Moran’s I value indicates stronger spatial autocorrelation40. ArcGIS 10.7 is used to obtain the Moran’s I index, expected index (E(I)), Z-score, and P-value for China’s digital industry. The Moran’s I index ranges from − 1 to 1, where a Moran’s I index greater than 0 indicates positive spatial autocorrelation. Moreover, if the Z-score exceeds 2.56 and the P-value is below 0.01, it indicates significant spatial clustering. Driving factors With the continuous growth of China’s digital industry, it has gradually exhibited characteristics such as being technology-intensive, highly permeable, uncertain, pioneering, and strategic, garnering significant attention. The digital industry has become a focal point of today’s economy, with increasing importance in technological innovation, economic growth, and social change. The development of the digital industry is influenced by various factors. We analyze these factors from three dimensions: infrastructure, industry development, and innovation potential (as shown in Table 1). Infrastructure: Key factors include broadband internet penetration, mobile phone penetration, and the number of relevant industrial parks. Broadband internet penetration measures the level of digitization, mobile phone penetration reflects the reach of digital services, and the number of relevant parks indicates the supportive environment for digital industry development41. These factors directly impact the spatial distribution of the digital industry. Industry Development: This includes internet-related output, inclusive development of digital finance, and software development.Internet-related output measures industry scale and contribution, digital finance inclusion gauges the accessibility of financial services, and software development indicates the vitality of the innovation ecosystem. These indicators illustrate the influence of industry scale and innovation capabilities on spatial distribution. Innovation Potential: Key factors are the number of patents granted and foreign investment. Patents measure innovation vitality, while foreign investment reflects the region’s attractiveness and competitiveness. These factors are used to evaluate the region’s innovation potential and appeal. By selecting these representative and explanatory factors, we can more accurately identify the true drivers of spatial dynamics, effectively avoiding interference from endogeneity issues. We use Geo Detectors and Geographically Weighted Regression (GWR) to explore the factors influencing the spatial distribution of China’s digital industry. Table 1Index system of location driving factors of digital industry in China.TypeInfluencing factorCodeThe specific meaning of the detection factorInfrastructureInternet broadband penetration rateX1Number of Internet broadband access users (ten thousand households)Mobile phone penetration rateX2Number of mobile phone users at the end of the year (ten thousand households)The number of industrial parksX3Number of national development zones (home)Industrial developmentInternet-related outputX4Per capita post and telecommunications business income (RMB / person)Inclusive development of digital financeX5Peking University Digital Financial Inclusion IndexSoftware developmentX6Number of software copyright registration scoreInnovation potentialPatent authorizationsX7The number of patents granted is scoredForeign investmentX8Attract foreign investment to score points Index system of location driving factors of digital industry in China. The Geo Detector is a spatial analysis model used to investigate the relationship between a geographic attribute and explanatory factors without making excessive assumptions, making it widely employed in research within both natural sciences and socio-economic impact studies42. Geo Detectors measures how well different factors explain spatial variation by comparing the level of spatial heterogeneity with and without those factors. In the context of multidimensional analysis of the spatial distribution and development of China’s digital industry across different cities, and considering the availability of actual data, we have utilized the factor detector module within the Geo Detector to select influencing factors. This approach aims to uncover the spatial differentiation patterns of China’s digital industry and the driving forces behind them In the formula, N represents the total number of sample units in the region, σ2 denotes variance, Nh and represent the sample size and variance of the hth class of influencing factors, and L signifies the number of classifications for the hth class of influencing factors. Furthermore, q represents the magnitude of the detection index, with an increasing value indicating a greater impact of the corresponding detection factor. Using ArcGIS 10.7 software and applying the natural break method, the driving factors were categorized into five levels and converted into categorical variables. Subsequently, the Geo Detector model was used to measure the influence magnitude of each factor on the spatial distribution of China’s digital industry. Geographically Weighted Regression (GWR) is an extension of linear regression that incorporates spatial structure into regression models to study spatial non-stationarity among variables from a local perspective43. GWR performs local regression at each location to capture spatially varying relationships between predictors and outcomes. In this study, the GWR model is utilized to estimate the varying impacts of different influencing factors on the number of publicly listed digital industry enterprises in different locations. This quantitative analysis accounts for the spatial heterogeneity of these impacts. The structure of the GWR model is as follows: Where (ui, vi) represents the geographical coordinates of the i-th sample spatial unit, and βk is the estimate for neighboring spatial observations, which varies with changes in location. Results Temporal evolution characteristics of China’s digital industry The patterns of industrial development are determined by both industry models and the dual characteristics of international and domestic environments. Analyzing the stages of development in China’s digital industry helps deepen our understanding of the spatial pattern, evolution, and influencing factors of the digital industry44. This study classifies the development stages of China’s digital industry from 1980 to 2020 into four periods based on statistical analysis of the number and growth rates of digital enterprises: the Early Exploration Period (1980–1990), the Initial Development Period (1991–2000), the Rapid Fluctuation Period (2001–2008), and the Dynamic Stability Period (2009–2020), as depicted in Fig. 2. Additionally, ArcGIS 10.7 software was used to apply the natural breaks method for classification and visualization, as shown in Fig. 3. Stage and number of listed enterprises in China’s digital industry in 1986–2020. Spatial distribution pattern of Listed enterprises in China’s digital industry. Early Exploration Period (1980–1990): The development of computer hardware technology laid the foundation for the digital industry, and China witnessed the establishment of its first batch of manufacturing-oriented digital industry enterprises, marking the beginning of the early exploration phase of the digital industry. During this stage, China’s digital industry started relatively early, but the number of publicly listed digital enterprises was relatively low, and they were concentrated in only 32 cities, primarily in coastal and economically developed regions. Manufacturing-oriented digital industry enterprises dominated this phase, accounting for approximately 75.76% of the total, while the number of service-oriented digital industry enterprises was relatively limited. Initial Development Period (1991–2000): The rise of semiconductor storage and semiconductor industries drove the initial development of the digital industry. Chinese companies began to recognize the potential of the digital industry and actively engaged in research and innovation. The digital industry began to expand nationwide, with cities across the country establishing digital industry enterprises, leading to the widespread development of the digital economy. During this period, the number of publicly listed service-oriented digital industry enterprises grew rapidly, achieving a 26-fold increase, and playing a significant role in China’s economic transformation and upgrading. Rapid Fluctuation Period (2001–2008): China’s accession to the World Trade Organization (WTO) presented development opportunities for the digital industry, marking a golden period for the core sectors of the digital economy. Structural adjustments and upgrades drove the rapid development of the digital industry nationwide. Cities across the country established digital industry enterprises, contributing significantly to China’s economic transformation and upgrading. During this period, the number of publicly listed service-oriented digital industry enterprises experienced substantial growth, further enhancing their role in China’s economic landscape. Dynamic Stability Period (2009–2020): The global financial crisis and fluctuations in international markets affected the digital industry, but the Chinese government implemented measures to promote digital trade, expanding the international market for the digital industry. With the emergence of new technologies, the digital industry entered a new phase of digital transformation and development within the digital economy. During this period, the central and western regions attracted a large number of digital industry enterprises and investments, leading to an increase in the number of publicly listed digital industry enterprises. The number of service-oriented digital industry enterprises increased significantly, accounting for approximately 80.72% of the total, gradually taking the lead in development. In the four periods of the digital industry, changes in the rate of entry of innovators and market concentration align with Schumpeter’s theory12 of “creative destruction.” During the Early Exploration Period (1980–1990), the number of innovators was relatively low due to the nascent state of technology. As technology advanced and market demand grew, the rate of entry of innovators significantly increased during the Initial Development Period (1991–2000). In the Rapid Fluctuation Period (2001–2008), the number of innovators surged dramatically, intensifying industry competition and challenging traditional enterprises. By the Dynamic Stability Period (2009–2020), key innovators became more stable, and market competition matured. Geographically, the digital industry was relatively dispersed in the early stages but saw the emergence of concentrated digital industry clusters in the mature stage. These changes reflect Schumpeter’s theory, illustrating how innovation drives industry concentration and market evolution. Based on the development stages of China’s digital industry, data from four time points in 1990, 2000, 2008, and 2020 were selected. The NNI and Standard Deviation Ellipse methods were used to explore the temporal evolution patterns of China’s digital industry as a whole and within the manufacturing and service sectors from 1990 to 2020, as shown in Table 2; Fig. 4. Table 2Analysis of the nearest neighbor and standard deviation elliptical parameters.TypeYearNNIZ-scoreCenterXE (°)CenterYN (°)xstdDist(km)ystdDist(km)Rotation (°)Area(km2)China’s digital industry19900.37-9.83117.3630.571095.87555.7829.741,913,426.6820000.19-49.29116.0930.291019.82680.9623.732,181,699.8420080.16-89.42116.1129.501031.79609.0722.081,974,278.4120200.17-106.75115.4529.981050.15692.8524.172,285,811.59service digital industry19900.55-6.05117.9930.371034.94429.4724.321,396,361.5420000.26-33.08116.7429.75960.56587.8123.731,773,827.3220080.21-59.00116.0428.13985.00508.6023.711,573,846.8120200.27-40.86115.7228.40941.44575.7623.461,702,879.86manufacturing digital industry19900.38-4.72115.3831.181282.52695.0040.972,800,263.0120000.20-32.21115.2830.981084.95757.9525.742,583,450.5620080.13-65.67116.1830.811033.41681.6723.422,213,077.9620200.16-97.13115.3830.361066.4713.8925.452,391,670.28 Analysis of the nearest neighbor and standard deviation elliptical parameters. Center of gravity and standard deviation ellipse of distribution. For the overall spatial distribution of China’s digital industry, when analyzing the average nearest neighbor, it exhibited significant spatial clustering characteristics. However, over time, the degree of clustering in China’s digital industry gradually decreased, showing a trend of gradual diffusion.Regarding the standard deviation ellipses, the clustering dynamics of China’s digital industry from 1990 to 2020 remained relatively stable, maintaining a distribution pattern that was generally “southwest (slightly west) to northeast (slightly east)” and moving slightly towards the southwest. This pattern is consistent with China’s socio-economic development orientation. The diminishing orientation of the standard deviation ellipse suggests an increasing level of industry clustering over this period, with industry elements tending to form increasingly tight spatial agglomerations. However, a comparison of the Nearest Neighbor Index (NNI) between the manufacturing and service sectors of the digital industry revealed some disparities. At different time points, the NNI for the manufacturing sector consistently exceeded those for the service sector, signifying a more pronounced clustering phenomenon in the spatial distribution of manufacturing digital industries when compared to their service sector counterparts. The service sector exhibited a broader distribution range, resulting in a more dispersed spatial pattern. Furthermore, based on observations of the Standard Deviation Wllipse, notable disparities in spatial dispersion and Center Point shifts emerged between the manufacturing and service sectors of the digital industry. The standard deviation ellipse’s area and radius for the service sector consistently surpassed those of the manufacturing sector, while the manufacturing sector experienced more pronounced shifts and changes in its centroid compared to the service sector. Spatial distribution characteristics of China’s digital industry Using kernel density analysis to visualize the spatial distribution characteristics of China’s digital industry, as shown in Fig. 5, the research findings suggest that China’s digital industry has exhibited a pattern of expansion from core cities to surrounding regions across different time periods, gradually evolving into a multi-center development model. Kernel density distribution of China’s digital industry from 1990 to 2020. From the perspective of different time periods: Early Exploration Period (1980–1990): During this era, China’s digital industry primarily thrived in core urban centers. High-density zones were predominantly concentrated in the Yangtze River Delta and the Pearl River Delta regions, forming a spatial structure characterized by a “dual-center” configuration centered around Shenzhen and the Shanghai-Suzhou axis. Initial Development Period (1991–2000): China’s digital industry continued its upward trajectory, with the Yangtze River Delta and the Pearl River Delta regions retaining their status as high-density areas. Additionally, the Beijing-Tianjin-Hebei region emerged as a high-density zone, shaping a spatial dynamic marked by a “tri-core” structure, with the “Yangtze River Delta-Pearl River Delta-Beijing-Tianjin-Hebei” axis at its core. Rapid Fluctuation Period (2001–2008): This period witnessed significant fluctuations in China’s digital industry. Coastal areas expanded further, while western regions began to experience gradual development. Consequently, a spatial configuration emerged characterized by “coastal areas in contiguous distribution and central cities scattered in distribution.” High-density areas in coastal regions extended to encompass Fujian Province, establishing a contiguous development corridor stretching from the Shandong Peninsula to the Yangtze River Delta, the western bank of the Taiwan Strait, and onward to the Pearl River Delta. Dynamic Stability Period (2009–2020): China’s digital industry transitioned into a phase of dynamic stability, manifesting a pattern of comprehensive expansion from central cities to outlying regions. The central and western regions exhibited a noticeable increase in kernel density, with numerous second and third-tier cities rapidly emerging as pivotal nodes in the digital industry landscape. From the perspective of both manufacturing and service digital industries: Manufacturing Digital Industry encompasses computer, communication, and other electronic equipment manufacturing, along with electrical machinery and equipment manufacturing. In 1990, publicly listed enterprises in computer, communication, and other electronic equipment manufacturing, as well as electrical machinery and equipment manufacturing, constituted roughly 82% of the total manufacturing sector. General equipment manufacturing only began to emerge in 1992. Subsequently, computer, communication, and other electronic equipment manufacturing, as well as general equipment manufacturing, embarked on a trajectory of rapid growth, experiencing year-on-year increases, while electrical machinery and equipment manufacturing saw a gradual decline. Service Digital Industries encompass software and information technology services, professional technical services, business services, wholesale, internet, and related services, among others. Among these, software and information technology services take the lead, with publicly listed enterprises in this sector accounting for over 43% of the total service sector across different time periods, with this proportion continuing to rise. Additionally, professional technical services, business services, internet wholesale and retail, and related services collectively constitute approximately 37% of the total service sector. In contrast, sectors such as news and publishing, monetary and financial services, motor vehicles, electronic products, and daily product repair have relatively smaller proportions of publicly listed enterprises within the total service sector. In summary, China’s digital industry distribution pattern can be described as a “one-core, three-poles, and multiple strong areas” configuration. Beijing serves as the core city, while Shanghai, Shenzhen, Hangzhou, and others act as engines of growth, playing pivotal roles in propelling the digital industry’s advancement. Beijing, renowned for its advanced information transmission, software, and information technology services, exerts a powerful influence through its radiation and leadership. These related enterprises also provide a strong impetus for the development of the digital industry in key cities such as Guangzhou, Suzhou, Xi’an, Nanjing, Chongqing, Chengdu, and Tianjin. In turn, Shanghai, Shenzhen, and Hangzhou form the “three poles,” with each leveraging its unique strengths to drive digital industry progress. Based on the distribution of publicly listed digital industry enterprises in various regions, Moran’s I index (Fig. 6), expected index E(I), Z-score, and P-value for China’s digital industry were obtained for the four time cross-sections in 1990, 2000, 2008, and 2020 using global spatial autocorrelation analysis. Trend plot of Moran’s I values of China’s digital industry from 1990 to 2020. Overall, the spatial distribution pattern of China’s digital industry has transitioned from a dispersed trend to a notably concentrated trend. Before 1990, the digital industry in China predominantly exhibited a dispersed pattern. From 1991 to 2000, it still showed some degree of dispersion. However, from 2001 to 2008, there was a significant shift in the spatial distribution trend, beginning to exhibit a certain degree of clustering. From 2009 to 2020, the trend toward industry clustering further intensified. Within the digital industry, there is a notable difference between the manufacturing and service sectors in terms of their spatial distribution trends over different years. Manufacturing digital industries gradually exhibited a clear trend toward clustering, whereas service digital industries have largely maintained a dispersed trend, albeit with some slight variations. For manufacturing digital industries, prior to 1990, there was noticeable dispersion in their spatial distribution. From 1991 to 2000, there was a faint emergence of spatial clustering, but it was not very significant. Between 2001 and 2008, manufacturing digital industries began to exhibit clear spatial clustering, although it was not overwhelmingly pronounced. On the other hand, for service digital industries, prior to 1990, there was a certain degree of dispersion in their spatial distribution, although it was not very prominent. From 1991 to 2000, service digital industries continued to exhibit dispersion, but the trend was weaker compared to the 1990s. Between 2001 and 2008, service digital industries still displayed a dispersed spatial trend, albeit even weaker. From 2008 to 2020, service digital industries began to show a faint trend toward clustering in spatial distribution, but this clustering was not very pronounced. This divergence in spatial trends between manufacturing and service digital industries likely reflects their distinct development paths influenced by technology, market dynamics, and policy factors. Manufacturing digital industries gradually formed pronounced clusters driven by technological advancements and favorable policies, while service digital industries, due to their inherent nature of dispersion, experienced limited clustering tendencies. Based on the previous analysis of spatial distribution using kernel density estimation and spatial autocorrelation in China’s digital industry, it is evident that a discernible spatial pattern exists in the clustering of digital industry clusters in China. The multi-center agglomeration of industries further enhances the international competitiveness of China’s manufacturing digital industry clusters. For the manufacturing digital industry, a continuous agglomeration area spanning administrative regions has emerged, primarily focusing on urban clusters in the middle and lower reaches of the Yangtze River and the middle and lower reaches of the Yellow River. In the urban cluster of the middle and lower reaches of the Yangtze River, anchored by core cities such as Suzhou and Hangzhou and with the synergy of surrounding cities like Wuhan, a concentration zone for electrical machinery and equipment manufacturing has formed. An exemplary cluster in this region is the Hangzhou digital security industry cluster, situated on the shores of West Lake and renowned as “China’s Silicon Valley.” Hangzhou has become a leading city in the digital security industry, with prominent companies like Dahua Technology and Hikvision leading the way, creating a dynamic digital security industry cluster. Furthermore, “China Optics Valley” in Wuhan is also noteworthy, as it gathers numerous optoelectronic information industry enterprises, including prominent firms like Yangtze Optical Fiber and Huagong Tech, aggregating innovative forces within the optoelectronics sector. In the urban cluster of the middle and lower reaches of the Yellow River, with the Shandong Peninsula urban cluster at its core, an agglomeration zone for general equipment manufacturing has been established. This region, with its abundant resources and strategic location, has attracted a substantial concentration of general equipment manufacturing enterprises. Through rational industrial layout and innovation-driven development, this concentration zone has gradually formed a regional industrial and value chain, providing strong support for the development of the digital industry. For the manufacturing digital industry, it is primarily concentrated in first-tier cities such as Beijing, Shanghai, and Shenzhen, which continue to be the primary agglomeration areas for the service digital industry. While some level of concentration is observed in more cities for the service digital industry, its degree of agglomeration remains relatively limited compared to the manufacturing sector. The spatial distribution of the service digital industry is still led by first-tier cities, with weaker diffusion trends. Taking the Hefei Intelligent Voice Industry Cluster as an example, it has leveraged the development of iFlytek to create “China’s Voice Valley.” With rapid growth in the field of intelligent voice technology, this cluster has demonstrated the potential for a service digital industry cluster to quickly thrive within a specific technological and innovative direction, becoming one of the important centers nationwide. Factors influencing the spatial distribution of China’s digital industry According to the Geo Detector results (Table 3), all eight driving factors under the three dimensions of infrastructure, industrial development, and innovation potential have a significant influence on the spatial distribution of China’s digital industry at the 0.01 level. Table 3Results of the influence factor detection.TypeDetection factorX1X2X3X4X5X6X7X8China’s digital industryq statistic0.2110.3170.3800.1800.4930.5240.5090.461p value0.0000.0000.0000.0000.0000.0000.0000.000manufacturedigital industryq statistic0.1330.2370.4360.1520.4120.3950.4430.335p value0.0000.0000.0000.0000.0000.0000.0000.000service digital industryq statistic0.2080.2890.5060.2150.2480.5290.4080.374p value0.0000.0000.0000.0000.0000.0000.0000.000 Results of the influence factor detection. manufacture digital industry Regarding the overall development of the digital industry, the primary core impact factors include software development and patent authorization, while the sub-core impact factors include the penetration rate of mobile phones, the number of industrial parks, digital finance inclusive development, and foreign investment. Other impact factors include internet-related output and Internet broadband penetration rate. When examining manufacturing and service digital industries, some differences emerge. Manufacturing-based digital industries place a greater emphasis on technological innovation, foreign investment, and intellectual property protection. In contrast, service-oriented digital industries rely more on software development capabilities, industrial parks, technological innovation, and international cooperation. The combined impact of these factors shapes the geographical distribution of manufacturing-based and service-oriented digital industries, providing unique support for the development of different industry types. Using ArcGIS 10.7 for a comprehensive regression analysis of influencing factors, the OLS regression model produced R2 = 0.594, Adjusted R2 = 0.579, and AICc = 632.023. Conversely, the GWR model yielded R2 = 0.646, Adjusted R2 = 0.615, and AICc = 698.454. Based on the R2 values, it is evident that the simulated outcomes of the GWR model are more logical than those of the OLS model. If the difference between the AICc values of the GWR model and the OLS model exceeds 3, it signifies that the GWR model’s fitting performance surpasses that of the OLS model25. In this study, the disparity between the two is approximately 66, underscoring the superior fitting performance of the GWR model. Taking into consideration the significance characteristics of various factors and the magnitude of their impact, we selected four indicators that passed the 99.9% significance test and had a median regression coefficient greater than 0.1. Using the GWR model, we conducted a detailed analysis of their spatial differentiation characteristics as influencing factors. These factors include the number of related industrial parks, the development of digital financial inclusion, software development, and foreign investment (Fig. 7). Spatial differentiation patterns of factors affecting the distribution. For the number of related industrial parks (Fig. 7a), the magnitude of its influence gradually decreases from the central region towards both sides, with some prefecture-level cities in the Yangtze River Delta having the most significant positive impact. The quantity of related industrial parks reflects the level of government attention to promoting industrial development. Local governments create a conducive environment for innovation and development by establishing and supporting these industrial parks. The Suzhou Industrial Park in China serves as a successful example, providing a platform for numerous technology and manufacturing companies to converge innovation resources, fostering industry clustering and collaborative innovation. Examining the development of digital financial inclusion (Fig. 7b), the influence on enterprise locations gradually decreases from south to north, with a weaker reliance on the northern regions for businesses’ location choices. Digital financial inclusion offers more financial support and financing channels, particularly for startups and innovative enterprises, driving innovation and development in the digital industry and attracting more participants. From a software development perspective (Fig. 7c), the overall impact on enterprise location choices shows a trend of decreasing influence from southeast to northwest, with a stronger positive effect on enterprise location choices in the southwest direction. Software development serves as the cornerstone of the digital industry, crucial for innovation and technological advancement, playing a driving role in the digital economy. In terms of foreign investment (Fig. 7d), the overall impact on enterprise location choices decreases from the southwest to the northeast, with the highest values along the coastal areas of the Pearl River Delta. This suggests that the Pearl River Delta region is more attractive to foreign investment, while the northeast region has relatively lower appeal for foreign investment. Foreign investment brings resources and networks that drive the transnational expansion of the digital industry. Discussion and conclusion In the previous section, we systematically analyzed the evolution and spatial patterns of China’s digital industry. Here, we integrate the discussion and conclusion to provide a cohesive reflection on the study’s theoretical and practical implications, in comparison to existing research. Compared to prior studies, our research offers a comprehensive nationwide perspective. While works such as Mao45 on the digital industry in the Yangtze River Economic Belt and Wu46 on the digital economy in Hangzhou primarily focus on specific regions, offering regional insights. In contrast, our study spans the entire country, deepening the understanding of national digital industry dynamics and highlighting both regional disparities and overarching development trends. Furthermore, earlier research by Song26 and Gu28 concentrated on the specific sub-sector clustering phenomena. Our research, however, not only differentiates between manufacturing and service sectors within the digital industry but also analyzes their respective clustering patterns. We discovered that the digital manufacturing sector emphasizes technological and production capabilities, while the service sector is more focused on software development and international collaboration. Our study systematically analyzes the temporal evolution, spatial distribution, and influencing factors of China’s digital industry, revealing four key stages of its development and presenting a spatial distribution pattern characterized by “one core, three poles, and multiple nodes.” The digital industry is shaped by a diverse set of factors, whose impacts vary across regions, underscoring the complexity of its growth. It fills a gap in the literature by presenting a comprehensive national perspective. It not only distinguishes between the manufacturing and service sectors but also delves into the spatial clustering and influencing factors unique to each. These findings contribute both theoretically and empirically to the understanding of the spatiotemporal dynamics of China’s digital industry, providing policymakers with the tools to formulate more precise regional development and industry upgrading strategies. Nevertheless, the study is not without limitations. The analysis is constrained to publicly listed companies, and the exploration of influencing factors lacks the depth required for a full understanding. Future research should aim to expand the data sources and incorporate qualitative methods, such as in-depth interviews and case studies, to achieve a more comprehensive and nuanced understanding of the digital industry’s evolution. This will further enhance the precision of policy-making and industry planning. Based on a comprehensive examination of the factors influencing the spatial distribution of the digital industry, this study offers the following policy recommendations to provide targeted guidance for government departments and decision-makers. First, accelerate the development of key digital infrastructure, including 5G networks, intelligent transportation systems, and data centers, with a focus on expanding coverage in rural and remote areas. Second, promote regional collaboration among digital industry clusters to foster cross-regional innovation and technology transfer within industrial chains. Third, improve the data governance framework to ensure data security and protect intellectual property rights. Fourth, implement targeted investment and financing support policies, especially for emerging technology sectors such as cross-border e-commerce and Artificial Intelligence Generated Content (AIGC). Finally, establish a global talent recruitment mechanism along with a local digital skills training system to ensure sustained innovation and competitiveness in the digital industry. These recommendations offer valuable guidance for government officials and decision-makers to support the industry’s long-term success. In conclusion, this study provides valuable insights into the evolution and spatial distribution of China’s digital industry, offering a robust reference for policy formulation and practical applications. As the digital economy continues to grow, sustained research will be essential for guiding its healthy development and enhancing China’s global competitiveness. Acknowledgements This work was supported by National Natural Science Foundation of China [41771136], Natural Science Foundation of Fujian Province [2022J01163], Natural Science Foundation of Heilongjiang Province, China [LH2019D009], Postgraduate Innovation Project of Harbin Normal University [HSDSSCX2022-09]. Author contributions Methodology, Data analysis, Writing-original draft: H.Z.; Supervision, Funding acquisition: S.W.,L.J.; Supervision, Resources, Project administration: F.L.; Writing-review , Project administration: J.X; Supervision: T.Z. Data availability The datasets generated and/or analysed during the current study are not publicly available due the data also forms part of an ongoing study, but are available from the corresponding author on reasonable request. Code Availability Codes for data analysis are available upon request. Declarations Competing interests The authors declare no competing interests. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Contributor Information Suqiong Wei, Email: suqiongwei@126.com. Lili Jiang, Email: jll0707@163.com. References Associated Data Data Availability Statement The datasets generated and/or analysed during the current study are not publicly available due the data also forms part of an ongoing study, but are available from the corresponding author on reasonable request. Codes for data analysis are available upon request.