Network Evolution of Digital Technology Transfers and Implications for Urban Digital Innovation Governance: Evidence from Chinese Patent Transactions

Abstract

Digital technology transfer plays a pivotal role in reshaping innovation landscapes and fueling the growth of the digital economy. To investigate this phenomenon, this study draws on data on digital technology transfers from the China National Intellectual Property Administration (CNIPA). Using tools such as Gephi 0.10.1 and ArcGIS 10.8, we construct an inter-city digital technology transfer network and develop a quantitative model to analyse the mechanisms by which it impacts urban digital innovation across multiple geographic scales. The main findings are as follows: (1) The inter-city digital technology transfer network in China forms a “diamond-shaped” spatial structure centred on Beijing, Shanghai, Guangzhou, and Shenzhen, with several regional hubs sustaining its connectivity and organisation. (2) Despite a decline in the proportion of intra-city transfers, the number of participating cities continues to rise, revealing a spatial pattern of diffusion from core cities toward inland provincial capitals. (3) Benchmark regression results show that both inter- and intra-city transfers significantly enhance urban digital innovation capacity, with inter-city transfers exhibiting a more substantial effect than their intra-city counterparts. This finding holds after a series of robustness tests. (4) Heterogeneity analysis, based on categorising cities into higher-tier (municipalities, sub-provincial cities, and provincial capitals) and lower-tier groups, indicates that the effect of digital technology transfer on innovation is more pronounced in lower-tier cities.

1. Introduction

The Fourth Industrial Revolution is advancing rapidly, with digital technology emerging as the core engine driving high-quality economic development [1]. Society is accelerating toward a profound integration of digital and physical spaces, signalling the full arrival of the digital technology era [2]. To secure a leading position in future development, governments at all levels are actively formulating digital technology innovation policies, committed to unleashing new productive forces and building new competitive advantages. In 2023, the Central Committee of the Communist Party of China and the State Council issued the “Overall Plan for Building a Digital China,” explicitly stating that “building a Digital China is a vital engine for advancing Chinese-style modernisation in the digital era.” It further identifies “establishing an independent and self-reliant digital technology innovation system” as the cornerstone of Digital China development. Innovation activities often occur within specific spatial contexts [3]. Similarly, digital technology innovation relies on the aggregation and reconfiguration of innovation factors—such as talent, capital, and technology—within particular spaces to generate sustained innovative momentum [4]. Digital technologies, including artificial intelligence, the Internet of Things, and 5G, drive cross-city collaboration among innovation entities through technology transfer, forming transregional digital technology transfer networks. From the perspective of regional innovation system theory, digital technology transfer between cities is embedded within innovation networks. Through network externalities, it promotes the aggregation and restructuring of innovation factors, transcending administrative boundaries and geographical distances to drive the spatial reconfiguration and functional optimisation of innovation systems [5]. This process not only strengthens the interactive connections among innovation entities within a region but also provides new impetus for enhancing cities’ digital technology innovation capabilities. Therefore, understanding the relationship between inter-city digital technology transfer and urban digital innovation holds significant importance.

In the field of technology transfer, it is regarded as a mechanism for optimising the flow of regional innovation resources and promoting coordinated regional development. Research methods primarily employ social network analysis [6,7,8] and multidimensional proximity [9,10], focusing on the analysis of network attributes, spatial characteristics, and influencing factors of knowledge transfer. Regarding urban innovation development, research emphasises comparative studies of regional innovation capabilities and spatial disparity analysis. Significant disparities in innovation efficiency exist across regions, with regional R&D activities exhibiting distinct characteristics. Methods such as the SPDM model [11] and fixed-effects models have been employed to analyse in depth the core factors influencing urban technological innovation capabilities [12]. As research has advanced, scholars have begun exploring the relationship between urban technology transfer and urban innovation. Li Qixiang et al. constructed and analysed a 166-year patent technology transfer network in the United States, revealing that the knowledge spillover effects of core cities and the absorption capacity of peripheral cities jointly form the key pathway for enhancing regional innovation capabilities. Within this framework, intra-city knowledge flow and recombination dominate the innovation system. At the same time, inter-city technology transfer plays a pivotal role in facilitating the exchange, integration, and diffusion of heterogeneous knowledge [13]. Related empirical studies also reveal that cities deeply embedded in technology transfer networks exhibit consistently higher innovation outputs, indicating technology transfer’s positive role in enhancing urban innovation capability [14,15]. Inter-city connectivity and cross-city technology transfer prove particularly effective in transmitting heterogeneous, complex knowledge, thereby boosting urban innovation performance [16,17]. Given the dominant role of digital technology in the new wave of technological revolution, the relationship between digital technology and urban digital innovation capacity has not received sufficient attention. Existing research focuses on how knowledge spillovers at different scales—local or non-local—impact urban innovation, but a single geographic scale struggles to fully explain the sustained effects of technology transfer on cities. As a multi-level process connecting local and non-local innovation activities, technology transfer exhibits significant variations in its impact on urban technological innovation capacity across different geographic scales. Therefore, delving into the transfer mechanisms of digital technologies within multi-scale spatial contexts and their role in urban digital innovation holds not only significant theoretical importance but also substantial practical value for optimising regional innovation layouts and promoting high-quality development of the digital economy.

Digital technology in the new economic landscape encounters constraints from deglobalization, unilateralism, and technological nationalism, limiting its widespread adoption [18]. International digital technology transfer now encounters multiple barriers. As a major player in digital technology, China must accelerate domestic economic circulation and deepen cross-regional digital technology transfer. This represents a crucial response to evolving international conditions and a forward-looking strategy to solidify the foundation of China’s digital economy and enhance its independent innovation capabilities. Therefore, strengthening China’s digital technology development to foster an efficient and collaborative domestic technology transfer system has become a critical research topic. By systematising and theorising the successful experiences and solutions China has developed in digital technology application, governance, and cross-regional transfer, such research can contribute “Chinese wisdom” to address the technology transfer challenges currently faced by developing countries worldwide

2. Theoretical Foundations and Research Hypotheses

In technology transfer activities, innovation agents promote cross-regional transfer of patented technologies by exchanging and sharing technical knowledge [19]. Innovation stems from interactions and collaborations among actors. Lundvall’s (1992) National Innovation System theory demonstrates that innovation is a multi-actor, systematic, and dynamic behavioural process [20]. Under national institutional arrangements and organisational frameworks, multiple actors form innovation networks that create conditions for knowledge generation, dissemination, and application. This, in turn, drives substantial innovation and enhances national competitiveness. Based on this, technology transfer is defined as the systematic transfer of technological elements across organisations and regions within an institutionalised framework for sharing. Its core purpose is to transform inventions into products and services that benefit society, thereby narrowing the gap between technological advancement and economic growth [21]. As the core spatial carriers of global technological competition [22], cities are not only vital venues for technology transfer but also strategic frontiers where nations vie for dominance in the new wave of technological revolution and industrial transformation. As network society theory reveals, cities construct transregional innovation networks through information and knowledge flows, thereby generating “spatial fluidity” [23]. Within this networked structure, technology transfer exhibits spatial unevenness, primarily manifested as the gradient transfer of technological elements from cities with high innovation potential to those with lower potential through knowledge spillover effects [24]. For technology recipients, technology transfer effectively reduces knowledge complexity, enabling integration and innovation upon existing knowledge foundations to accelerate technological catch-up. For technology exporters, it optimises the spatial layout of innovation and enhances overall competitiveness. Therefore, this paper proposes the following hypothesis:

Hypothesis 1.

Urban digital technology transfer enhances a city’s digital technological innovation capabilities.

Digital technologies, including cloud computing, the Internet of Things, and artificial intelligence, are widely applied in R&D, manufacturing, and other processes. They effectively overcome issues such as information fragmentation, departmentalisation, and asymmetry, improve the efficiency of supply-demand matching, and facilitate the flow of production factors among various innovation entities [25]. Due to differences in geographical location and resource endowments, the transfer of digital technologies leads to variations in digital innovation across cities. On the one hand, enterprises serve as the primary agents of technology transfer in China. By leveraging emerging internet technologies to deepen industrial digital transformation, they generate multiplier effects for economic growth and comprehensively elevate total factor productivity. Digital transformation has thus become a critical lever for enterprises transitioning from scale expansion to quality enhancement [26]. However, China’s high-tech enterprises exhibit uneven distribution, being highly concentrated in economically robust, transportation-accessible, and market-rich regions such as the eastern coastal areas, inland provincial capitals, and municipalities directly under the central government [24]. On the other hand, the theory of absorptive capacity indicates that an organisation’s or region’s innovation performance depends not only on the supply of external knowledge but also on its ability to identify, absorb, and utilise this knowledge [27]. High-innovation cities control the diffusion and application rights of innovation resources within digital technology transfer networks formed among cities. As their innovation levels continue to rise, the “disruptive” technologies available through local technology transfer become increasingly limited, hindering further enhancement of their innovation capabilities. In contrast, for cities with lower innovation levels, resource redundancy and overload do not impede their ability to enhance innovation through technology transfer. They are more likely to achieve technological catch-up through external knowledge. Thus, we propose the following hypotheses:

Hypothesis 2.

The impact of inter-city digital technology transfer on digital technology innovation capabilities exhibits regional heterogeneity across cities.

Hypothesis 3.

Inter-city technology transfer exerts a more substantial promotional effect on a city’s digital technology innovation capabilities than intra-city digital technology transfer.

3. Research Methods and Data Sources

3.1. Data Sources and Processing

3.1.1. Data Recognition

Data is sourced from the China National Intellectual Property Administration (CNIPA) and collected via web scraping for patent transfer records from 2000 to 2024. Retrieved data includes multiple metrics such as patent title, IPC classification number, applicant address before transfer, applicant address after transfer, postal code, legal status, and registration effective date. During data processing and analysis, multiple assignees were identified within the legal status field. Following established practices, Python 3.5 was used to split P1→P2P3 into P1→P2 and P1→P3 (where P2 and P3 represent the transferor and assignee entities, respectively, and → indicates the transfer direction).

The data was processed into a dataset containing registration effective date, IPC patent numbers, transfer-out location, and transfer-in location. Based on the “Key Digital Technology Patent Classification System (2023),” which encompasses seven patent classification systems—artificial intelligence, high-end chips, quantum information, Internet of Things, blockchain, industrial internet, and metaverse—IPC patent numbers were extracted from these systems. These include G06F3*, G06F8*, G06F9*, G06F11*, G06F12*, G06F13*, G06F15*, G06F16*, G06F17*, G06F21*, G06F30*, where the asterisk (*) denotes inclusion of all classification numbers at that level and below in the International Patent Classification. We filtered patent transfer data from 2000 to 2024 and categorised them into intra-city and inter-city technology transfers based on whether the transfer’s origin and destination locations were identical. Inter-city technology transfer occurs when patent technology originates in one city and is received by another. In contrast, intra-city technology transfer occurs when the technology originates from and is received within the same city.

3.1.2. Phases of Digital Technology Transfer

National policy incentives are closely linked to technology transfer. Since China acceded to the WTO in 2001, the opening of foreign markets and domestic independent innovation have synergised, driving a rapid surge in patent numbers. Based on the Implementation Plan for the National Technology Transfer Promotion Action, the Law of the People’s Republic of China on the Transformation of Scientific and Technological Achievements, the National Technology Transfer System Construction Plan, the Science and Technology Progress Law of the People’s Republic of China, the Notice on the National Technology Transfer System Construction Plan, the Statistical Classification of the Digital Economy and Its Core Industries (2021), and the Key Digital Technology Patent Classification System (2023), As shown in Figure 1, China’s digital technology transer is categorized into three phases: the embryonic stage (2000–2010), Formative Stage (2011–2018), Growth Stage (2019–2021), and Stabilization Stage (2022–2024).

3.2. Variable Specification and Econometric Model Design

3.2.1. Variable Declaration

(1) Dependent Variable: Digital Patents: This study measures digital technological innovation capability using the number of digital technology patents granted at the city level. Compared to the comprehensive concept of “urban innovation capability,” existing multi-indicator systems often suffer from double-counting and difficulties in aligning measurement criteria, making it challenging to reflect all levels [28]. In the digital economy era, digital technologies—represented by artificial intelligence, blockchain, cloud computing, and big data—continue to push beyond traditional boundaries, accelerating the intelligent transformation of cities. These technologies deeply integrate the real and digital economies, serving as the core force driving China’s high-quality economic development. Patents are the direct result of innovation and a significant outcome of innovative output. Given their early and widespread use in scientific innovation, patents were ultimately selected. City-level digital economy patent authorisation data from the CNRDS database, covering 2001 to 2023, was to represent the level of urban digital innovation.

(2) Core Explanatory Variables: This study selects the number of patents for technology transfers within cities and between cities as core explanatory variables.

3.2.2. Model Construction

The dependent variable is lagged. Let $Digita{l}_{t+2}$ denote the number of digital patent grants for an individual $i$ in year $t+2$, ultimately expressed in tens of thousands. $In{t}_{i,t}$ (intra-city technology transfer), $Ou{t}_{i,t}$ (inter-city technology transfer) $Ct{r}_{i,t}$ represents a series of control variables, as shown in

Table 1. Description of Related Variables.

	Variable Name	Measurement	Data Source
Dependent variable	Urban Digital Technology Innovation Capability	Digital Economy Patent Authorisations	CNRDS Database
Independent variables	Inter-city Digital Technology Transfer	Transfer of Digital Patent Technologies Between Prefecture-Level Cities	State Intellectual Property Office
	Digital Technology Transfer Within the City	Digital Patent Technology Transfer Within Prefecture-Level Cities	State Intellectual Property Office
Control variables	Level of economic development	GDP per capita	City Statistical Yearbook
	Human capital	Number of Full-time Faculty Members at Regular Higher Education Institutions	City Statistical Yearbook
	Foundations of Higher Education	Number of employees in comprehensive scientific research technical services	City Statistical Yearbook
		Scientific Expenditure	City Statistical Yearbook
	Share of Fixed Asset Investment	Share of Fixed Asset Investment	City Statistical Yearbook
	Technology R&D	Research and Development Personnel	City Statistical Yearbook

, ${\gamma }_{i,t}$ and ${\lambda }_{i.t}$ denotes cities and year fixed effects, $\alpha$ is the constant term, ${\beta }_1$ and ${\beta }_2$ are the fitting coefficients, and ${\epsilon }_{i,t}$ is the random disturbance term. The model incorporates city- and year-fixed effects and employs city-clustered robust standard errors to address heteroskedasticity and intra-group autocorrelation simultaneously.

$$D{\mathit{igital}}_{i,t+2}=\alpha +{\beta }_{1}Ou{t}_{i,t}+{\beta }_{2}Ct{r}_{i,t}+{\gamma }_{i,t}+{\lambda }_{i.t}+{\epsilon }_{i,t}$$

$$D{\mathit{igital}}_{i,t+2}=\alpha +{\beta }_{1}I{\mathit{nt}}_{i,t}+{\beta }_{2}Ct{r}_{i,t}+{\gamma }_{i,t}+{\lambda }_{i.t}+{\epsilon }_{i,t}$$

4. Digital Technology Transfers Across Different Geographical Scales

4.1. Spatio-Temporal Evolution of Inter-City Digital Technology Transfer

Within China’s inter-city digital technology transfer network, network density reflects the concentration of digital technology transfers; average path length and average clustering coefficient indicate the presence of a “small-world” network; network diameter represents the maximum distance between any two nodes; and average weighted degree reflects the strength of node connections.

As shown in

Table 2. Structural Characteristics of Inter-city Digital Technology Transfer Networks, 2000–2024.

	Emergence Stage	Formative Stage	Growth Stage	Stable Stage
Visualization
Network Density	0.011	0.093	0.147	0.225
Average Degree	3.957	35.788	54.72	83.575
Average Weighted Degree	35.473	658.827	1123	1518.054
Average Clustering Coefficient	0.573	0.683	0.703	0.693
Average Path Length	2.595	1.964	1.857	1.736
Network Diameter	6	4	3	3

, this paper uses Gephi to analyse the various stages of the technology transfer network. From 2000 to 2024, China experienced growth in inter-city digital technology transfer, which contributes to the continued advancement of digital technologies. The network visualisation reveals a transition from a sparse to a denser, more complex network. Network density increased markedly from 0.011 to 0.255, reflecting increasingly frequent interregional technological exchanges. The average clustering coefficient increased from 0.573 to 0.693, while the average clustering coefficient length decreased from 2.5595 to 1.736. The increase in average path length, coupled with the decrease in average clustering coefficient, indicates a “small-world” property within China’s digital technology transfer network. Both the average degree and the average weighted degree showed substantial growth, with the average degree rising from 3.957 to 83.575 and the average weighted degree increasing from 35.473 to 1518.054. This indicates continuous growth in the number of nodes within the digital technology transfer network, alongside a significant improvement in the quality of technological interactions between nodes [28]. Cities are increasingly interconnected while simultaneously exhibiting a degree of “core-periphery” differentiation in their structural patterns.

In the initial stage, metrics such as the number of edges, network density, average degree, and network diameter remained relatively low. The overall connectivity of inter-city digital technology transfer was weak, with most cities facing the dilemma of “no transfer” or “difficult transfer” of digital technologies. During the formation period, the connectivity of inter-city digital technology transfer improved, with significant increases in metrics like the number of edges, network density, average degree, and diameter. During the growth phase, the network structure deepens further. As the scale of technology transfer continues to expand, the average path length decreases to 1.857. This suggests that the growth in high-speed rail connectivity between cities and improvements in internet technology are associated with shorter spatiotemporal distances and reduced transportation and information costs for technology transfer. On average, approximately 1.8 steps are required to facilitate digital technology transfer interactions between cities, with the longest path not exceeding 3.000, consistent with the “small-world” phenomenon. During the growth phase, regional clustering in digital technology transfer between cities moderated, decreasing from 0.703 in 2019–2022 (growth phase) to 0.693. This indicates that in the new phase, digital transfers between cities are no longer confined to urban clusters like “Beijing-Tianjin-Hebei,” “Yangtze River Delta,” “Pearl River Delta,” and “Guangdong-Hong Kong-Macao.” Increasingly, cities outside these clusters are participating in digital technology transfer activities.

PageRank indices and proximity centrality were calculated in Gephi for the Emergence, Formation, Growth, and Stabilisation phases to observe changes in the status and functions of nodes within the digital technology transfer network. Due to the large number of cities, the top 20 cities are listed in

Table 3. 2000–2024 PageRank and Proximity Centrality.

	Emergence Stage		Formative Stage		Growth Stage		Stable Stage
	PageRank	Closeness	PageRank	Closeness	PageRank	Closeness	PageRank	Closeness
1	Beijing	Beijing	Beijing	Beijing	Beijing	Beijing	Guangzhou	Guangzhou
2	Shenzhen	Shenzhen	Shenzhen	Shenzhen	Shenzhen	Shenzhen	Shenzhen	Shenzhen
3	Shanghai	Shanghai	Guangzhou	Guangzhou	Shanghai	Shanghai	Beijing	Beijing
4	Guangzhou	Guangzhou	Shanghai	Shanghai	Chengdu	Guangzhou	Shanghai	Shanghai
5	Hangzhou	Nanjing	Chengdu	Chengdu	Guangzhou	Chengdu	Suzhou	Suzhou
6	Chengdu	Hangzhou	Suzhou	Suzhou	Hangzhou	Hangzhou	Xi’an	Xi’an
7	Nanjing	Suzhou	Tianjin	Dongguan	Quanzhou	Shaoxing	Hefei	Hefei
8	Wuhan	Dongguan	Dongguan	Tianjin	Shaoxing	Suzhou	Nanjing	Chengdu
9	Suzhou	Chengdu	Quanzhou	Quanzhou	Suzhou	Quanzhou	Chengdu	Nanjing
10	Xi’an	Tianjin	Chongqing	Chongqing	Nanjing	Nanjing	Hangzhou	Hangzhou
11	Tianjin	Wuhan	Shaoxing	Shaoxing	Hefei	Hefei	Chongqing	Chongqing
12	Jinan	Wenzhou	Ningbo	Ningbo	Wuhan	Wenzhou	Wuhan	Wuhan
13	Dalian	Wuxi	Hangzhou	Hangzhou	Wenzhou	Wuhan	Tianjin	Tianjin
14	Dongguan	Changchun	Wuxi	Nanjing	Tianjin	Dongguan	Quanzhou	Jinan
15	Wenzhou	Xi’an	Nanjing	Wuxi	Xi’an	Xi’an	Jinan	Quanzhou
16	Wuxi	Shenyang	Nantong	Nantong	Dongguan	Tianjin	Wuxi	Wuxi
17	Shenyang	Changzhou	Taizhou	Taizhou	Taizhou	Taizhou	Harbin	Harbin
18	Changchun	Jinan	Wuhan	Foshan	Qingdao	Chongqing	Xuzhou	Dongguan
19	Changsha	Dalian	Foshan	Qingdao	Zhengzhou	Zhengzhou	Dongguan	Xuzhou
20	Ningbo	Ningbo	Qingdao	Wenzhou	Chongqing	Xuzhou	Foshan	Foshan

. The PageRank index provides a more comprehensive measure of the centrality of individual nodes within China’s urban digital technology transfer network [29]. Closeness centrality indicates a node’s ability to avoid control by other nodes. Nodes with high closeness centrality can efficiently acquire information and rapidly disseminate it throughout the network [30]. Based on PageRank and proximity centrality rankings, Beijing, Shenzhen, Shanghai, and Guangzhou consistently ranked among the top cities from 2001 to 2024. These cities are the core of the digital technology transfer network and significant sources of digital technology, confirming their unshakable “structural power” within the digital technology ecosystem. The “status” of highly influential city nodes within the digital technology transfer network remained essentially unchanged across different periods. Leveraging their formidable research capabilities, economic strength, and abundant talent pools, the “Beijing-Shanghai-Guangzhou-Shenzhen” quartet exerts significant radiating and leading influence within the established “diamond-shaped” network structure of technology transfer. As shown in Figure 2, with the rapid advancement of digital technologies, other cities have progressively enhanced their digital infrastructure, leading some digital technologies to converge in industrial powerhouses like Suzhou, Xi’an, and Hefei. Nevertheless, the eastern coastal regions, characterised by higher levels of industrial development, remain the core of the digital technology transfer network.

4.2. Spatial-Temporal Evolution of Technology Transfer in Urban Areas

As shown in Figure 2, between 2000 and 2004, cities experiencing large-scale intra-city digital technology transfers were highly concentrated in eastern coastal urban clusters, including Beijing-Tianjin-Hebei, the Yangtze River Delta, Chengdu-Chongqing, and the Pearl River Delta. As shown in

Table 4. Ranking of Digital Technology Transfers Hubs Within the City.

	Emergence Stage	Formative Stage	Growth Stage	Stable Stage
1	Beijing	Beijing	Beijing	Beijing
2	Shenzhen	Shenzhen	Shenzhen	Shenzhen
3	Shanghai	Shanghai	Guangzhou	Shanghai
4	Hangzhou	Guangzhou	Shanghai	Guangzhou
5	Guangzhou	Hangzhou	Suzhou	Hangzhou
6	Nanjing	Suzhou	Hangzhou	Suzhou
7	Suzhou	Nanjing	Nanjing	Nanjing
8	Shenyang	Chengdu	Dongguan	Wuhan
9	Wuhan	Dongguan	Jinan	Chengdu
10	Tianjin	Wuhan	Wuhan	Chongqing
11	Chongqing	Wuxi	Chengdu	Xi’an
12	Chengdu	Tianjin	Chongqing	Tianjin
13	Dongguan	Zhengzhou	Qingdao	Jinan
14	Xi’an	Chongqing	Xi’an	Hefei
15	Ningbo	Ningbo	Tianjin	Qingdao
16	Xiamen	Changzhou	Wuxi	Dongguan
17	Qingdao	Foshan	Changsha	Wuxi
18	Jiaxing	Qingdao	Nanchang	Changsha
19	Wuxi	Xi’an	Harbin	Changzhou
20	Jinan	Jinan	Hefei	Xuzhou

, Inland regions saw transfer activities primarily centred in provincial capitals, such as Zhengzhou, Hefei, and Jinan. From 2000 to 2010, 25,081 digital technology patents underwent ownership transfers across 217 cities. While eastern coastal cities exhibited substantial intra-city transfers, cities in central, western, and northeastern regions also showed significant volumes of these transfers. For instance, Shenyang and Jinan recorded 507 and 214 intra-city digital technology transfers, respectively, during this period. From a city-wide perspective, 54 cities recorded over 50 digital technology transfers during this period. Beijing, Shenzhen, Shanghai, Hangzhou, and Guangzhou ranked 1st to 5th nationally with 5438, 4083, 3316, 849, and 763 transfers, respectively. From 2011 to 2017, intra-city digital technology transfers increased to 284,031 cases, though their share of total transfers declined to 69%. The number of cities engaging in intra-city digital technology transfers rose to 311, with 59 cities exceeding 500 cases each. Among these, Beijing, Shenzhen, and Guangzhou each recorded more than 10,000 intra-city digital technology transfers. During the 2019–2021 phase, the proportion of intra-city digital technology transfers further declined to 55%. A total of 315 cities engaged in intra-city digital transfers during this period, with provincial capitals and municipalities directly under the central government conducting more such activities internally. Beijing, Shanghai, Shenzhen, and Guangzhou continued to maintain their leading positions. Between 2022 and 2024, Beijing, Shenzhen, and Guangzhou all experienced a decline in the scale of intra-city digital technology transfers compared to the period from 2019 to 2021. Conversely, Shanghai and Hangzhou, both in the top five, showed a significant upward trend, indicating continued expansion in the scale of digital transfers. The proportion of intra-city digital technology transfers dropped to 52%, with 326 cities participating. Overall, while the scale of digital technology transfers within core cities decreased, more cities engaged in intra-city digital technology transfer activities.

The ongoing technological revolution is associated with growth in productivity and changes in the global innovation landscape. Innovation leadership has transitioned from a few traditional centres to a multipolar diffusion pattern. Emerging high-tech industries display a spatial distribution characterised by “high coastal concentration—gradient inland adoption—polarised urban cluster-driven growth,” spanning multiple provinces. Because technological innovation in emerging high-tech industries relies on heterogeneous information, internal information flows within a single city, and interactions grounded in local embeddedness and face-to-face communication, these may not fully meet higher-level innovation requirements.

5. The Impact of Digital Technology Transfers on Urban Digital Innovation Capabilities

The depiction of spatio-temporal differentiation patterns in digital technology transfer not only illuminates the phenomenon itself but also offers key insights into the underlying pathways that shape urban innovation capability. Specifically, technology transfer demonstrates pronounced spatial concentration and evolving network characteristics over time. Building on this observation, this study proposes that digital technology transfer systematically transforms a city’s digital innovation capability through multiple parallel mechanisms. To test this proposition, the analysis will begin by examining the overall effect using a benchmark regression model, followed by the sequential identification and verification of specific pathways through correlation analysis.

5.1. Baseline Regression

The VIF value of 2.03 is well below 10, indicating no multicollinearity issues among the variables. The corresponding regression coefficients (R²) are 0.827 and 0.876, indicating that the model fits the data well. As shown in

Table 5. Baseline regression results.

Variable	(1)	(2)
Intra-city Technology Transfer	1.234 **
	(3.01)
Inter-city Technology Transfer		4.455 ***
		(16.41)
Time Fixed	Yes	Yes
City Fixed	Yes	Yes
Sample Size	6153	6153
R2	0.8270	0.8760

Notes: **, *** indicate p < 0.05, p < 0.01; Heteroscedasticity-robust standard errors in parentheses.

, the estimated coefficients for the core explanatory variables are all positive and statistically significant at the 5% level. The analysis indicates that both intra-city and inter-city digital technology transfers are associated with enhanced digital technology innovation capabilities within cities, consistent with Hypothesis 1. According to the regression results, the coefficient for inter-city digital technology transfer (4.455) is greater than that for intra-city digital technology transfer (1.234). Compared to intra-city digital technology transfer, inter-city digital technology transfer has a more significant effect on enhancing a city’s digital innovation capacity, thereby validating Hypothesis 3.

Possible reasons are as follows: The new round of technological revolution has advanced productivity and continuously reshaped the global innovation landscape [31]. Innovation leadership has shifted from a few traditional powerhouse cities to a more multipolar pattern of diffusion. Emerging high-tech industries exhibit a spatial pattern characterised by “high concentration along the coast—gradient adoption inland—polarised driving by urban clusters,” which is widely distributed across various provinces. Given that technological innovation in emerging high-tech industries heavily relies on heterogeneous information, the information supply within a single city, along with interactions based on local embeddedness and face-to-face communication, struggles to meet higher-level innovation demands and may even act as an obstacle [32].

5.2. Stability Test

Given the unique characteristics of municipalities directly under the central government and the significant differences in digital technology innovation capabilities between cities like Beijing, Shanghai, Guangzhou, and Shenzhen, and other cities, this study conducted a regression analysis. The analysis was performed after excluding specific samples to test the robustness of the conclusions. The results are presented in

Table 6. Robustness Test Results.

	Exclude Certain Samples		One-Period Lag		Two-Period Lag
Variable	(1)	(2)	(3)	(4)	(5)	(6)
Intra-city Technology Transfer	1.258 *		4.036 *		1.258 *
	(2.69)		(2.23)		(2.69)
Inter-city Technology Transfer		4.346 ***		5.835 ***		4.346 ***
		(11.49)		(3.85)		(11.49)
Time Fixed	Yes	Yes	Yes	Yes	Yes	Yes
City Fixed	Yes	Yes	Yes	Yes	Yes	Yes
Sample Size	6069	6069	6446	6446	6153	6153
R2	0.7780	0.7793	0.7255	0.8132	0.6778	0.7708

Notes: *, *** indicate p < 0.1, p < 0.01; Heteroscedasticity-robust standard errors in parentheses.

, Column (1) and Column (2): After excluding municipalities directly under the central government, urban digital technology innovation capability exerts a significant positive influence, confirming Hypothesis 1. The potential mechanism lies in the following: Cities with relatively lagging overall development, under pressure to achieve high-quality growth, enhance their innovation capability through cross-regional flows of technological factors. This enables “technological catch-up,” becoming a key pathway to boost digital technology innovation capability and subsequently drive high-quality development of the digital economy.

The regression coefficient for inter-city digital technology transfer exceeds that of intra-city transfer, confirming Hypothesis 3. This stems from the “combination-based” nature of digital innovation, where technological interactions across heterogeneous cities facilitate the cross-reconfiguration of innovation resources [33]. Concurrently, knowledge gradients and disparities in infrastructure and talent endowments between cities amplify knowledge exchange and spillover effects, thereby further catalysing digital technological innovation.

Table 6 presents the robustness test results for the number of digital patent grants, lagged by one and two periods, as the dependent variable. The regression structure indicates that inter-city digital technology transfer exerts a more substantial promotional effect on a city’s digital technology innovation capability than intra-city digital technology transfer. Consistent with the benchmark regression results, this further substantiates this study’s conclusions.

5.3. Heterogeneity Test

Resource availability and policy regulations, such as limitations in R&D talent, funding, and institutional support, constrain the effectiveness of technology transfer. This results in varying impacts of technology transfer on urban innovation capability across different administrative tiers of cities [12]. Drawing on relevant research, cities are categorised into higher-tier cities (including municipalities directly under the central government, sub-provincial cities, and provincial capitals) and lower-tier cities. As shown in

Table 7. Results of the Regional Heterogeneity Test.

	High-Level Cities		Low-Level Cities
Variable	(1)	(2)	(3)	(4)
Intra-city Technology Transfer	1.258 ***		4.036 ***
	(3.50)		(3.73)
Inter-city Technology Transfer		4.346 ***		5.835 ***
		(5.17)		(6.58)
Time Fixed	Yes	Yes	Yes	Yes
City Fixed	Yes	Yes	Yes	Yes
Sample Size	756	756	5397	5397
R2	0.8341	0.8846	0.7121	0.7217

Notes: *** indicate p < 0.01; Heteroscedasticity-robust standard errors in parentheses.

, the regression coefficients for both high-level and low-level cities are positive and statistically significant at the 1% level. However, the coefficient for low-level cities is higher than that for high-level cities, indicating that the impact of digital technology transfer on urban digital technology innovation capabilities is heterogeneous across administrative levels, thus supporting Hypothesis 2. In terms of regression coefficients, lower-tier cities exhibit significantly higher coefficients than higher-tier cities. This stems from higher-tier cities generally possessing more advanced industrial foundations, more robust technological infrastructure, higher levels of informatisation, and denser concentrations of universities and research institutions. These factors result in higher overall levels of digital technology innovation and stronger innovation capabilities. To further elevate their digital technology innovation capabilities, cities require the inflow of “disruptive” technologies. Given the limited local information and resources, they should actively engage with internationally renowned innovation hubs, such as those in New York and Tokyo.

6. Conclusions and Discussion

6.1. Conclusions

This paper uses IPC patent numbers referenced in the “Classification System for Key Digital Technology Patents (2023)” to screen data and construct a digital technology transfer network spanning 2000–2024. It conducts an in-depth exploration of the evolutionary characteristics of this digital technology transfer network. Due to data availability constraints, the CNRDS database was selected to obtain the number of authorised digital economy inventions from 2001 to 2023, representing cities’ digital technology innovation capabilities. This study investigates whether digital technology transfer can effectively enhance cities’ digital technology innovation capabilities. The main conclusions are as follows:

From 2000 to 2024, China’s digital technology transfer experienced rapid development. Key characteristics include a continuous rise in the inter-city technology transfer share, while the intra-city transfer share declined steadily yet still accounted for 50% of total digital technology transfer. Evolutionary features are summarised from both inter-city and intra-city dimensions:

First, inter-city digital technology transfer characteristics: (1) The inter-city digital technology transfer network has become denser, with transfer efficiency continuously improving. The “distance” of digital technology interaction between cities has steadily shortened, exhibiting a “small-world” phenomenon within the network. (2) The evolution of China’s inter-city digital technology transfer network exhibits regional clustering, forming a diamond-shaped structure centred on core hubs in the Beijing-Tianjin-Hebei region, Pearl River Delta, Yangtze River Delta, Sichuan-Chongqing, and middle-lower Yangtze River urban clusters. Polarisation is increasingly pronounced, with eastern coastal regions becoming key drivers of domestic digital technology transfer while western regions continue to face challenges in transferring digital technologies. (3) The number of city nodes participating in digital technology transfer continues to grow. Central China plays an irreplaceable intermediary role in digital technology transfer, with its geographical position serving as a bridge between eastern and western regions.

Second, characteristics of intra-city digital technology transfer: (1) China’s intra-city digital technology transfer activities are highly concentrated in eastern coastal regions such as the Beijing-Tianjin-Hebei region, the Yangtze River Delta, and the Pearl River Delta, which serve as growth poles. With the deepening of market-oriented reforms and the urgent demand for open innovation, the constraints imposed by urban administrative boundaries on the spatial flow of digital technologies in China have begun to weaken. Cross-city digital technology flows have become the new norm, and a unified national digital technology trading market is gradually taking shape. (2) Cities like Beijing, Shanghai, and Shenzhen, which previously exhibited large-scale intra-city digital technology transfers due to their abundant talent pools and advantageous locations, remain at the forefront of national intra-city technology transfers despite a slight reduction in scale. (3) The number of cities participating in intra-city digital technology transfers has increased.

Third, this study reveals changes in the scale of digital technology transfer’s impact on urban innovation capability. While affirming the importance of both inter-city and intra-city digital technology transfer, it finds that the promotional effect of intra-city technology transfer on urban digital innovation has weakened. Based on regression analysis of digital technology transfer’s impact on urban digital innovation, the following conclusions are drawn: (1) Digital technology transfer enhances digital innovation capability; (2) Inter-city digital technology transfer exhibits a more substantial promotional effect than intra-city transfer; (3) The impact of digital technology transfer on urban innovation development is constrained by factors such as regional policies, leading to heterogeneity in cities’ digital technology innovation capabilities.

6.2. Discussion

At the national policy level, it is necessary to strengthen regional collaborative innovation and leverage the leading role of urban agglomerations. Currently, China’s digital technology transfer exhibits significant regional imbalances, necessitating concerted efforts to promote coordinated development across the eastern, central, western, and northeastern regions. It is recommended to establish regional counterpart support mechanisms through technology transfer and talent exchange, effectively linking the eastern areas with higher levels of innovation development to the central, western, and northeastern areas with relatively weaker innovation capabilities. This will promote the rational flow and optimal allocation of innovation factors nationwide, forming an innovation landscape characterised by complementary advantages and mutual advancement.

At the city level, efforts should be made to improve local policy systems and enhance innovation-driven effectiveness. (1) Strengthen the policy guidance role of local governments. Developed cities can draw on the experience of Hangzhou, which introduced the “Hangzhou Software and Information Services Industry Innovation and Development Three-Year Action Plan (2013–2015)” as early as 2012 to advance the digital economy sector systematically. In 2018, it further established the goal of becoming the “Number One City in the Digital Economy,” continuously refining its industrial ecosystem to evolve into a nationally renowned “Digital City gradually.” All regions should formulate sustainable, targeted digital industry promotion policies tailored to their specific circumstances. (2) Establish corrective mechanisms for technological innovation in underdeveloped regions. In areas experiencing market failures, governments must play a pivotal role: on the one hand, correcting potential innovation distortions stemming from overreliance on intellectual property protection; on the other, designing innovation incentive policies that fully account for constraints such as environmental regulations. By leveraging complementary effects across institutional frameworks, a systematic policy support system can be established to effectively stimulate digital technological innovation. (3) Promote a gradual development path for the digital economy. For moderately developed cities, it is recommended to first focus on cultivating and attracting leading digital economy enterprises to leverage their demonstration and driving effects fully. By supporting the growth of leading local enterprises, we gradually attract more digital industries to cluster locally and further enhance the appeal to outstanding external digital enterprises, forming a virtuous cycle.

This study still has certain limitations. While digital patent data can serve as an indicator for measuring a city’s digital innovation level, its representativeness remains insufficient. In reality, digital technology transfer between cities involves multiple factors such as talent mobility, government collaboration, cross-regional capital and goods flows, and investment spillover effects. Future empirical research requires deeper exploration. Subsequent studies should conduct industry- and sector-specific analyses to reveal the characteristics and hierarchical impacts of technology transfer at different spatial scales (e.g., international and interprovincial). Simultaneously, investigating the dynamic evolution and driving mechanisms of regional digital technology transfer networks from multiple dimensions—including innovation actors, technological structures, and spatial networks—will be a key direction for future research. Externalities generated by intra-urban technology transfer—such as adjustments in firm spatial layout, industrial upgrading, and shifts in talent structures—should also be incorporated into the research scope. Furthermore, patent licensing, which to some extent determines patent technology transfer, presents endogeneity issues. Future studies should develop a more comprehensive digital technology innovation indicator system to measure innovation capabilities while incorporating mechanisms and endogeneity tests into the model design.