Research on Evaluation and Influencing Factors of Regional Digital Innovation Ecosystem Resilience—Empirical Research Based on Panel Data of 30 Provinces and Cities in China

Abstract

The current global situation is complex and volatile. Thus, promoting the construction of a resilient digital innovation ecosystem has become an important issue for regional innovation development. Based on panel data from 30 Chinese provinces and cities, this study empirically investigated the current situation and resilience level of China’s regional digital innovation ecosystem development using a global entropy method, a technique for order performance by similarity to the ideal solution (TOPSIS), Moran’s I, obstacles, and panel models. The results indicate that the resilience level of China’s regional digital innovation ecosystem will grow from 2011 to 2021. Spatially, it showed a spatial distribution of “coastal high–central flat–northwest depression,” with inter-regional “low coupling” and intra-regional “high cohesion” characteristics. The global Moran’s I was greater than zero, decreasing by 43.860% from 2011 to 2021, and the scope of the “high-high” quadrant and the “low-high” quadrant has been expanding and narrowing, respectively. The state and response indicators were the main obstacles to the resilience development of the regional digital innovation ecosystem, and the obstacle degree of the pressure indicators increased during the survey period. Industrial structure, regional urbanization, human capital, and digital industrialization have made significant positive contributions to regional digital innovation ecosystem resilience overall, and the drivers were regionally heterogeneous in space. Finally, this study proposes strategies for improving the resilience of regional digital innovation ecosystems, including strengthening top-level design, differentiated development, and lowering obstacles.

1. Introduction

With the rapid development of digital technologies such as artificial intelligence, big data, and cloud computing, data have become a new production and innovation factor that promotes the transformation of traditional industries from “intellectual change” to “qualitative change” and has caused a deep integration of regional economic development and digital innovation. A digital innovation ecosystem is a socio-ecological system formed by industrial subjects related to digital innovation based on competitive cooperation [1]. Actors and business activities at different levels are linked by digitalization and form a complex network structure [2] involving various elements, including organization, technology, and knowledge [3] that is characterized by digitalized elements, virtualized subjects, and ecological relationships. With the rapid development of digital technology, digital innovation is driving the digital transformation of all fields, and the construction of a digital innovation ecosystem has become inevitable for high-quality regional development. However, as the world economy enters the volatility, uncertainty, complexity, and ambiguity (VUCA) era, the risk of international technology competition has risen sharply, posing serious challenges to the operation and development of regional digital innovation ecosystems. Thus, given the difficulty in predicting external shocks and disturbances, reducing the risks generated by external environmental factors and promoting the sustainable and healthy development of regional digital innovation ecosystems have become urgent issues. In this context, resilience, the ability of an economic or organizational system to absorb shocks, adapt to the environment, and recover rapidly [4], has gradually become a popular research topic in organizational management [5] and provides a new perspective for solving the above problems. Introducing the concept of resilience into the theory of digital innovation ecosystems, building resilient regional digital innovation ecosystems, and improving the resilience of systems to resist external shocks and perturbations have become key paths to promoting the healthy development of regional innovation systems. Current research on the resilience of digital innovation ecosystems mostly describes and analyzes the internal structure and operational processes of a single system. However, an overall evaluation and comparative study of multiple systems is lacking, which restricts our understanding of regional digital innovation ecosystems.

Therefore, this study considers the regional digital innovation ecosystem as the research object, measures the resilience of China’s regional digital innovation ecosystems to external shocks from 2011 to 2021 from a provincial perspective [6], analyzes the spatial and temporal evolution characteristics of the resilience level, identifies influencing factors, and proposes strategies to improve resilience. The innovative points of this study are as follows. First, we explain the concept of regional digital innovation ecosystem resilience, focusing on the system’s ability to resist external shocks and disturbances. This provides a scientific understanding of the current developmental status of system resilience, which is an extension of the digital innovation ecosystem theory. Second, by analyzing the connotations of the system resilience concept, we construct a resilience evaluation index system to scientifically evaluate the resilience level, which is an effective supplement to existing resilience research. Finally, we analyzed and discussed the spatial and temporal evolution of the resilience level breakthrough spatial limitations and attempted to fill the gap in the evolution of the spatial pattern of resilience. Simultaneously, identifying the barriers and drivers affecting the resilient development of regional digital innovation ecosystems can indicate the direction for the resilience construction of regional digital innovation ecosystems at this stage.

2. Literature Review

Joseph Schumpeter first introduced the concept of “innovation at the beginning of the twentieth century, which has since evolved into a variety of societal development paradigms. Lundvall proposed an “innovation system” from a systemic perspective. In 1992, Cooke [7] put forward the concept of a “regional innovation system”, and in the following year, Nelson suggested the concept of a “national innovation system”. Scholars’ understanding of the innovation process tends to be nonlinear, and innovation systems have become an important paradigm for innovation research. In the same year, Moore [8] proposed the concept of “business ecosystem” in the context of ecosystem theory in biology, arguing that a business ecosystem could be defined by an “economic association based on organizational interaction,” which pioneered the study of innovation ecosystem theory. In 2006, Adner [9] combined ecological theory with technological innovation and proposed the concept of an innovation ecosystem from the perspective of enterprises, asserting that innovation cannot be accomplished independently by individual enterprises and that innovation ecosystems should be built to strengthen inter-organizational collaboration and achieve value co-creation. At this point, the innovation paradigm shifted from traditional linear innovation 1.0 and coupled with interactive innovation 2.0 to an innovation ecosystem 3.0. An innovation ecosystem is an organic system formed by the interaction of heterogeneous organizational, institutional, economic, political, and technological elements. Collaborative symbiotic behaviors [10] and value relationships [11] exist among innovation subjects, and their symbiotic patterns and evolutionary processes have different characteristics at different stages of system development [12]. Innovation subjects intermingle with natural, social, economic, and other innovation environments to form complex networks of innovation ecosystems [13]. The external environment provides the conditions for communication and interaction between system subjects [14]. The innovation ecosystem has its own unique operation, growth, and evolution laws [15] and presents certain life cycle characteristics. One key inherent feature is its dynamic evolutionary nature [16], which emphasizes the coevolution among various types of innovation subjects [17].

With the rapid development of digital technology, digital transformation has penetrated the entire field of enterprise upgrading and social governance, expanding the theoretical boundaries of traditional innovation ecosystems and triggering research on digital innovation. Digital innovation refers to the development of a new product or service that combines physical products with digital technologies [18]. With the rapid progress in information technology and the continuous upgrading of digital devices, digital technology is deeply bound to physical components [19], giving rise to a new innovation paradigm (i.e., digital innovation) that significantly affects the rules of current operations in various organizations, industries, and fields [20]. Digital innovation is characterized by borderlessness, a focus on uncertainty-based development, and blurred boundaries between innovation processes and outcomes [21], all of which drive the digital transformation of innovation ecosystems. A digital innovation ecosystem (the combination of digital innovation and an innovation ecosystem) possesses not only the characteristics of a general innovation ecosystem (e.g., openness and complexity), but also the unique attributes of the digital era (e.g., platform, networking, modularity, and borderlessness). The digital innovation ecosystem is a new form of the traditional innovation ecosystem in the context of the digital era. Digital transformation promotes the digitization of innovation elements and data elementalization, which are continuously transferred and reorganized along with the generation, development, and elimination of digital technology; the shareable and reusable attributes of digital elements, in turn, promote the development of digital technology [22]. Digital technologies and elements mutually reinforce and drive the digital transformation of innovative ecosystems. In terms of system construction, scholars have mainly investigated the formation and evolution of the digital innovation ecosystem and how digitalization powers the transformation of traditional innovation to digital innovation. Beltagui [1] explored the impact of disruptive innovation on the traditional innovation ecology based on a 3D printing ecosystem and suggested that disruptive innovation helps force the system to adapt to the current development environment, achieve system change, and keep up with the times.

The word resilience is derived from the Latin word “resilio”, which means “to return to the original state before the event”. In 1973, Holling [23] introduced resilience to the field of ecology to describe the characteristics of natural ecosystems that maintain a stable state. Since then, scholars have gradually expanded the study of resilience to social sciences [24], which is reflected in urban planning and construction [25], community management [26], and regional economies. A social system is believed to be a nonequilibrium dynamic evolutionary process [27] that adjusts its structure and developmental state in response to external shocks. However, there is no specific equilibrium stable state [28] from which urban, economic, and innovation ecosystem resilience can be derived. Urban resilience is the response of urban systems to external shocks [29]. It is the product of the superposition of sustainable material systems and dynamic human social systems, in which the construction and development of human social systems is the key to, and foundation for, the role of material systems [30]. Economic resilience refers to the adaptive adjustment capacity of macroeconomic systems when subjected to external shocks [31]. It is considered an inherent property of a region and is not measured after a particular shock, as it constantly changes with the evolutionary development of the system itself and shifts in the external environment [32]. Roundy and Brockman [33] theorized the formation mechanism of entrepreneurial ecosystem resilience based on two dimensions: consistency and diversity. Meerow [29] considered resilience to be a system’s ability to self-learn, adapt, recover, and eventually achieve a new equilibrium state under external shocks. Wildavsky [34] proposed that the general system resilience consists of six elements: equilibrium force, compatibility, mobility, flatness, buffering, and redundancy. The measurement of system resilience has been the focus of resilience research and is mainly divided into two approaches: qualitative analysis and quantitative calculations. The former conducts descriptive research on the resilience of a specific system at the macro level, emphasizing the analysis of the overall composition and mechanism of action, whereas the latter seeks to accurately calculate system resilience using evaluation models and specific index systems. In terms of evaluation methods, the comprehensive index, functional model, threshold, social network, and scenario analysis methods are widely used, and research on the construction of resilience indicator systems is advancing. Schlör [35] suggested several urban resilience indicators, including productivity, infrastructure, quality of life, equity, and environmental sustainability. Bruneau [36] argued that a system’s functional level changes over time and constructed a system resilience evaluation model, thus providing a theoretical basis for relevant quantitative research.

In summary, with the continuous advancement of digital transformation in various societal fields, the innovation research paradigm has evolved from innovation systems and ecosystems to digital innovation ecosystems, triggering extensive academic discussions and research. Research on digital innovation ecosystems involves definitions of concepts, structures, and practical case analyses. Due to continuous improvements in the related theoretical analysis framework, regional digital innovation ecosystems have become an important research area. Simultaneously, with the changing development situation worldwide, the concept of resilience has gradually emerged as a research hotspot in the social sciences, giving rise to various resilience concepts (e.g., urban and economic resilience). In this context, studying the resilience of digital innovation ecosystems is particularly important. Therefore, we select the digital innovation ecosystem as the research object and empirically investigate the resilience of China’s regional digital innovation ecosystem from a provincial perspective. To this end, we constructed a resilience evaluation index system, measured the resilience level, discussed its spatial and temporal evolution processes, and identified barriers and driving factors. This study contributes to the existing literature in two ways. First, we define the connotation of digital innovation ecosystems based on the traditional innovation ecosystem theory, analyze the key role of digital innovation in the evolution of the innovation system, and extend the innovation theory in the context of the digital economy. Second, we propose the concept of regional digital innovation ecosystem resilience and establish a resilience evaluation index system, which provides a research tool for subsequent relevant empirical studies and expands the existing resilience theory. Thus, we hope to provide theoretical guidance for digital innovation practices and promote the construction of resilient regional digital innovation ecosystems.

3. Theoretical Framework and Research Methods

3.1. Theoretical Framework of Regional Digital Innovation Ecosystem Resilience

In this paper, we argue that external shocks faced by regional digital innovation ecosystems are characterized by non-linearity, suddenness, and unpredictability, which can seriously impact system functions. Such shocks are commonly observed during financial crises; thus, studying system resilience is particularly important. System resilience includes three dimensions: initial negative impact reduction (absorptive capacity), system adaptation (adaptive capacity), and post-impact recovery (resilience) [37]. Based on the above, the resilience process is decomposed into stages using the pressure-state-response model [38]: the results of which are shown in Figure 1. The PSR model is a causal framework for describing and analyzing interactions between society and the environment. It contains pressure, state, and response indicators, which indicate the pressure on the system, current state, and subjective initiative in response to a disaster. Accordingly, regional digital innovation ecosystem resilience is separated into three processes: “before the shock” (pressure), “during the shock” (state), and “after the shock” (response), corresponding to the aforementioned three indicators of the PSR model. Specifically, before a shock occurs, the system identifies potential risks and formulates a timely response plan to improve stability. A complex and robust system structure is a key characteristic of organizational resilience [39] that prevents system collapse due to insufficient buffering. When a shock occurs, the regional digital innovation ecosystem relies on its resilience to face it directly. On the one hand, heterogeneous innovation subjects flow across regions and promote convergent innovation by increasing the frequency of communication and interaction to cope with external shocks and disturbances. However, there are factor endowment differences among regional digital innovation ecosystems. The difference triggers the overflow of innovation factors such as information, talent, material, and funds [40]; promotes knowledge combination and radical innovation [41]; supports talent development processes, resource allocation, and technology trading; and enables the mobilization of resource elements in response to external shocks. Finally, after a shock occurs, all innovation subjects in the system carry out innovation activities to restore the regional digital innovation ecosystem functions and achieve the evolutionary development of the system. Under market selection, the innovation population evolves by changing its gene frequency to adapt to the changes in the external environment. This is manifested by the fusion of various innovation subjects; changing the form and structure of innovation factors through self-adaptation, self-learning, and self-adjustment; updating the innovation input–output paradigm to enhance innovation efficiency; realizing the spiral rise of the system function level at the spatial and temporal scales; and evolving to a higher form. From the shock perspective, the system remains stable and reduces the impact of external shocks and perturbations by responding to each stage of shock occurrence. Therefore, this study considers regional digital innovation ecosystem resilience as a multistage process in which the system uses its multidimensional capabilities, such as self-adaptation, self-learning, and self-adjustment, to cope with and resist shocks and enable the system function level to recover and develop smoothly. As the total volume of China’s economy continues to increase, the level of industrial structure is also rising, which provides good basic conditions for the development of regional digital innovation ecosystem resilience. However, at the same time, the problem of unbalanced and insufficient development should be noted, and there is a large gap in the development level between regions. Therefore, we propose the following hypotheses:

Hypothesis 1.

China’s overall regional digital innovation ecosystem resilience level has improved, but the difference in resilience levels between regions is significant.

3.2. Research Methods

3.2.1. Global Entropy Method

The traditional entropy method can only horizontally evaluate different regions in a year or vertically evaluate the resilience level of a region in different years, making the evaluation results incomparable. Therefore, we used the global entropy method to calculate the indicator weights. T data tables were arranged from top to bottom to form an mT × n specification global evaluation matrix, and the entropy method [42] was used to calculate the indicator weights using the following process. The specific steps were as follows:

$$\begin{array}{c}{P}_{ij}=\frac{u_{ij}^{’}}{\sum _{i=1}^{mT}{u}_{ij}^{’}}\end{array}$$

(1)

$$\begin{array}{c}{E}_j=-\frac{\sum _{i=1}^{mT}{P}_{ij}\mathit{ln}\left(P_{ij}\right)}{\mathit{ln}\left(mT\right)}\left(when P_{ij}=0,E_j=0\right)\end{array}$$

(2)

$$\begin{array}{c}{w}_j=\frac{1-E_j}{n-\sum _{j=1}^n{E}_j}\end{array}$$

(3)

where $u_{ij}^{’}$ denotes the standardized value, $P_{ij}$ denotes the proportion of the ith system in the index under the jth index, $E_j$ denotes the information entropy value of the jth index, and $w_j$ denotes the global weight value of each indicator.

3.2.2. TOPSIS Method

In the conventional TOPSIS method, the weight of each index is the same, by default. In this study, we used the results of the global entropy method as weights for each indicator to reduce the errors. We then used the TOPSIS method to measure the resilience level of regional digital innovation ecosystems. The steps are as follows [43].

$$\begin{array}{c}{A}_j^{+}=max\left\{u_{ij}^{’}∣i=\mathrm{1,2},3,\cdots ,m\right\}\end{array}$$

(4)

$$\begin{array}{c}{A}_j^{-}=min\left\{u_{ij}^{’}∣i=\mathrm{1,2},3,\cdots ,m\right\}\end{array}$$

(5)

$$D_i^{+}=\sqrt{{\sum }_{j=1}^n{w_j\ast \left(u_{ij}^{’}-A_j^{+}\right)}^2}\left(i=\mathrm{1,2},3,\cdots ,m\right)$$

(6)

$$\begin{array}{c}{D}_i^{-}=\sqrt{{\sum }_{j=1}^n{w_j\ast \left(u_{ij}^{\mathrm{’}}-A_j^{-}\right)}^2}\left(i=\mathrm{1,2},3,\cdots ,m\right)\end{array}$$

(7)

$$\begin{array}{c}{C}_i=\frac{D_i^{-}}{D_i^{-}+D_i^{+}}\end{array}$$

(8)

where $A_j^{+}$, $A_j^{-}$ denote the maximum and minimum values of the j-th index, respectively. $D_i^{+}$, $D_i^{-}$ denote the optimal solution and the worst solution, respectively. The closeness $C_i$ reflects the closeness of the evaluation object to the positive ideal solution; the larger the value, the higher the resilience level.

3.2.3. Moran’s I

We used Moran’s I to analyze the spatial correlation of the resilience level of regional digital innovation ecosystems. First, the spatial correlation characteristics of the resilience level were portrayed from an overall perspective, with the global Moran’s I taking a value range of [−1, 1]; the larger the absolute value, the stronger the spatial correlation. We then calculated the local Moran’s I and divided the global space into several regions to explain the spatial correlation of the resilience level from a microscopic perspective. The formula [44] for Moran’s I is as follows:

$$\mathrm{G}\mathrm{l}\mathrm{o}\mathrm{b}\mathrm{a}\mathrm{l} \mathrm{M}\mathrm{o}\mathrm{r}\mathrm{a}\mathrm{n}\mathrm{’}\mathrm{s} \mathrm{I}=\frac{{\sum }_{i=1}^n{\sum }_{j=1}^n{W}_{ij}(x_i-\stackrel{-}{x})(x_{\mathrm{j}}-\stackrel{-}{x})}{S^2{\sum }_{i=1}^n{\sum }_{j=1}^n{W}_{ij}}$$

(9)

$$\mathrm{L}\mathrm{o}\mathrm{c}\mathrm{a}\mathrm{l} \mathrm{M}\mathrm{o}\mathrm{r}\mathrm{a}\mathrm{n}\mathrm{’}\mathrm{s} \mathrm{I}=\frac{(x_i-\stackrel{-}{x}){\sum }_{j=1}^n{W}_{ij}(x_{\mathrm{j}}-\stackrel{-}{x})}{S^2}$$

(10)

where $n$ denotes the total number of regional digital innovation ecosystems; $W_{ij}$ denotes the weight value of each indicator; $x_i$ and $x_{\mathrm{j}}$ represent the observed values for provinces $i$ and $\mathrm{j}$, respectively. $\stackrel{-}{x}$, $S^2$ denote the mean and variance of $x_i$, respectively.

3.2.4. Obstacle Model

Clarifying the obstacles that affect the resilience development of regional digital innovation ecosystems and adopting targeted development strategies can improve resilience. We used an obstacle degree model [45] to calculate the obstacle degrees of various indicators and dimensions and then compared and analyzed the obstacle degrees of each stage.

$$I_{ij}=1-u_{ij}^{’}$$

(11)

$$G_j=W_j{P}_j$$

(12)

$$O_{ij}=\frac{I_{ij}{G}_j}{{\sum }_{j=1}^n{I}_{ij}{G}_j}\times 100\%$$

(13)

$$O_{iJ}=\sum _{j=J_1}^{J_k}{O}_{ij}$$

(14)

where $I_{ij}$ is the indicator deviation; $G_j$ is the contribution of the $j$th indicator to the dimension to which it belongs; $P_j$ is the weight of the $j$th indicator to the dimension to which it belongs; $O_{ij}$ is the obstacle degree of the $j$th indicator of the $i$th sample; $O_{iJ}$ is the obstacle degree of the $J$th dimension of the $i$th sample; $k$ is the number of indicators of the $J$th dimension.

4. Index System Construction

4.1. Construction of a Comprehensive Evaluation Index System

Drawing on relevant research results [46,47,48], a comprehensive evaluation index system was based on the principles of science and comprehensiveness, in which the pressure, state, and response dimension indicators reflected the potential risks faced by the regional digital innovation ecosystem, the system’s resilience to external shocks, and the system’s ability to autonomously recover after the impact (see

Table 1. Regional digital innovation ecosystem resilience evaluation index system.

Stage Layer	Indicator Layer	Unit
Pressure	P1: Social unemployment rate	%
	P2: Eco-stress index	-
	P3: External trade dependence	%
	P4: Level of government administrative monopoly	-
	P5: Unit R&D cost	-
State	S1: E-commerce penetration rate	%
	S2: Total telecommunications business	100 million CNY
	S3: GDP per capita	CNY
	S4: Technology contract turnover	100 million CNY
	S5: Total retail sales of social consumer goods	100 million CNY
	S6: Total number of patents granted	piece
	S7: Number of high quality scientific papers published	piece
	S8: Number of R&D topics in higher education	item
	S9: Number of graduates from general higher education institutions	person
	S10: Digital inclusive finance index	-
Response	R1: Digital industry R&D personnel equivalent full time equivalent	person-year
	R2: Internal expenditure on R&D funding for digital industry	100 million CNY
	R3: Growth rate of fixed asset investment in digital industry	%
	R4: Digital industry main business income	100 million CNY
	R5: Digital industry new product sales revenue	100 million CNY
	R6: Balance of deposits in financial institutions	100 million CNY
	R7: R&D investment intensity	%
	R8: Science and technology financial expenditure	100 million CNY
	R9: Number of internet broadband access subscribers	10,000 households
	R10: Fiber optic cable line length	kilometer

). Equations (15)–(17) explain these indicators.

$$\mathrm{P}2=\frac{{ECD}_i}{{GDPD}_i}$$

(15)

where ${ECD}_i$ and ${GDPD}_i$ denote the energy consumption of digital industry and gross domestic product of the digital industry, respectively.

$$\mathrm{P}4=\sqrt{\frac{{SOIE}_i}{{TIE}_i}\times \frac{{SOAI}_i}{{TAI}_i}}$$

(16)

where ${SOIE}_i$, ${TIE}_i$, ${SOAI}_i$, and ${TAI}_i$ denote the number of state-owned industrial enterprises above the designated size, number of industrial enterprises above the designated size, amount of investment in fixed assets of state-owned enterprises, and total investment in fixed assets, respectively.

$$\mathrm{P}5=\frac{{EONP}_i}{{NPSR}_i}$$

(17)

where ${EONP}_i$ and ${NPSR}_i$ denote expenditures on new product development and new product sales revenue of industrial enterprises above the designated size, respectively.

4.2. Data Source and Processing

Owing to delays in statistical data, we selected panel data from 30 provinces and cities in China from 2011 to 2021. The data sources were the China Statistical Yearbook, China Science and Technology Statistical Yearbook, China Torch Statistical Yearbook, China High Technology Industry Statistical Yearbook, and Statistical Yearbooks of Provinces and Cities. For example, we can directly obtain indicator data such as “GDP per capita” and “Technology contract turnover” for each region from the China Statistical Yearbook, while data such as “Eco-stress index” and “Unit R&D cost” need to be calculated on a region-by-region basis. Missing data were completed using the mean or interpolation method, and we finally obtained a total of 8250 data values for 30 regions in China from 2011 to 2021. The data were standardized, including 20 positive and 5 negative indicators, as follows [49]:

$$\begin{array}{c}{u}_{ij}^{’}=\frac{u_{ij}-u_{min}}{u_{max}-u_{min}}(\mathrm{Positive} \mathrm{indicator})\end{array}$$

(18)

$$\begin{array}{c}{u}_{ij}^{\mathrm{’}}=\frac{u_{max}-u_{ij}}{u_{max}-u_{min}}(\mathrm{Negative} \mathrm{indicator})\end{array}$$

(19)

where $u_{ij}$ denotes the $j$th indicator observation of the $i$th object and $u_{max}$ and $u_{min}$ denote the maximum and minimum observations of the indicator, respectively. $u_{ij}^{\mathrm{’}}$ is a standardized value.

5. Empirical Analysis of the Resilience Level of Regional Digital Innovation Ecosystems

5.1. Resilience Level Analysis of Regional Digital Innovation Ecosystems

5.1.1. Analysis of the Time Series Evolution of the Resilience Level

Table 2. Regional digital innovation ecosystem resilience level from 2011 to 2021.

Region	2011	2012	2013	2014	2015	2016	2017	2018	2019	2020	2021	Average
Beijing	0.247	0.263	0.287	0.308	0.338	0.357	0.382	0.416	0.453	0.486	0.517	0.368
Tianjin	0.120	0.149	0.170	0.181	0.191	0.195	0.190	0.201	0.200	0.215	0.233	0.186
Hebei	0.125	0.132	0.141	0.151	0.160	0.175	0.186	0.202	0.216	0.231	0.253	0.179
Shanxi	0.086	0.096	0.108	0.111	0.123	0.133	0.143	0.153	0.168	0.179	0.196	0.136
Inner Mongolia	0.092	0.098	0.106	0.115	0.120	0.129	0.132	0.142	0.150	0.162	0.175	0.129
Liaoning	0.131	0.145	0.157	0.165	0.170	0.168	0.176	0.186	0.195	0.206	0.218	0.174
Jilin	0.100	0.109	0.105	0.124	0.133	0.140	0.150	0.157	0.162	0.175	0.180	0.140
Heilongjiang	0.094	0.104	0.110	0.116	0.127	0.137	0.144	0.155	0.162	0.174	0.184	0.137
Shanghai	0.153	0.169	0.180	0.196	0.219	0.233	0.249	0.272	0.298	0.317	0.356	0.240
Jiangsu	0.232	0.259	0.282	0.303	0.335	0.360	0.383	0.423	0.473	0.533	0.609	0.381
Zhejiang	0.169	0.183	0.202	0.214	0.241	0.258	0.276	0.308	0.343	0.375	0.411	0.271
Anhui	0.119	0.128	0.141	0.154	0.171	0.184	0.195	0.213	0.240	0.258	0.289	0.190
Fujian	0.116	0.126	0.144	0.149	0.167	0.179	0.192	0.211	0.231	0.241	0.292	0.186
Jiangxi	0.112	0.118	0.124	0.132	0.154	0.154	0.168	0.183	0.202	0.215	0.235	0.163
Shandong	0.186	0.205	0.221	0.229	0.263	0.283	0.306	0.323	0.326	0.362	0.423	0.284
Henan	0.143	0.152	0.175	0.189	0.211	0.221	0.243	0.267	0.272	0.300	0.340	0.228
Hubei	0.139	0.148	0.164	0.182	0.201	0.210	0.226	0.252	0.281	0.290	0.338	0.221
Hunan	0.120	0.129	0.138	0.148	0.162	0.173	0.186	0.207	0.232	0.253	0.279	0.184
Guangdong	0.237	0.275	0.307	0.312	0.381	0.423	0.477	0.554	0.624	0.690	0.750	0.457
Guangxi	0.100	0.106	0.117	0.126	0.139	0.151	0.164	0.175	0.187	0.201	0.219	0.153
Hainan	0.106	0.105	0.111	0.125	0.148	0.158	0.158	0.160	0.169	0.174	0.180	0.145
Chongqing	0.096	0.106	0.114	0.125	0.149	0.162	0.182	0.193	0.214	0.223	0.244	0.164
Sichuan	0.156	0.133	0.147	0.160	0.187	0.202	0.217	0.246	0.269	0.292	0.320	0.212
Guizhou	0.100	0.088	0.097	0.109	0.130	0.139	0.149	0.159	0.172	0.184	0.191	0.138
Yunnan	0.079	0.088	0.096	0.105	0.121	0.146	0.147	0.159	0.177	0.193	0.204	0.138
Shaanxi	0.112	0.124	0.147	0.149	0.165	0.183	0.191	0.212	0.232	0.249	0.279	0.186
Gansu	0.082	0.094	0.108	0.118	0.131	0.140	0.137	0.148	0.156	0.167	0.179	0.133
Qinhai	0.073	0.094	0.091	0.130	0.134	0.125	0.130	0.137	0.152	0.163	0.179	0.128
Ningxia	0.075	0.085	0.095	0.125	0.120	0.128	0.133	0.142	0.150	0.158	0.169	0.126
Xinjiang	0.074	0.079	0.087	0.100	0.112	0.122	0.132	0.144	0.158	0.167	0.184	0.124
National average	0.126	0.136	0.149	0.162	0.180	0.192	0.205	0.223	0.242	0.261	0.288	0.197

and Figure 2 present the results of the calculation of the resilience levels of China’s regional digital innovation ecosystem from 2011 to 2021. On the whole, the spatial distribution of the resilience level shows significant inter-regional differences and the spatial distribution pattern of “coastal high–central flat–northwest depression” forms. Coastal regions generally have higher resilience levels and certain leading advantages, whereas inland regions have lower resilience levels and have not yet established regional synergies. During the survey period, the resilience level of the regional digital innovation ecosystem showed a trend of progressive development from the eastern coast to the inland regions, and the gap between the regions expanded. For example, Guangdong, Jiangsu, and other developed regions with resource advantages allow the level of resilience to grow rapidly, whereas Ningxia, Xinjiang, and other underdeveloped regions develop slowly, with the range of resilience levels increasing from 0.174 (2011) to 0.581 (2021). At the same time, the difference in resilience levels within each major economic zone is relatively small. For example, the cities of Shanghai, Jiangsu, and Zhejiang, which are situated within the Yangtze River Delta Economic Circle, have similar resilience levels and more balanced development. Thus, it can be seen that while the resilience level of China’s regional digital innovation ecosystem has improved and the overall development trend is stable, inter-regional “low coupling” and intra-regional “high cohesion” characteristics are prominent, and there is a certain degree of uneven development.

From the perspective of change, China’s regional digital innovation ecosystem resilience level has maintained steady growth from 0.126 (2011) to 0.288 (2021), with an average annual increase of 8.621%. The regional digital innovation ecosystem resilience levels increased by more than 100% from 2011 to 2021, with Guangdong having the largest increase of 216.248% and Liaoning lagging by only 66.292%. Changes in the level of resilience strongly correlated with the development situation. For example, on the one hand, with the “Belt and Road Digital Economy International Cooperation Initiative,” the “Outline Development Plan for the Guangdong-Hong Kong-Macao Greater Bay Area,” and a series of other policies, the resilience level of Guangdong increased from 0.423 (2016) to 0.750 (2021), with an average annual growth of 12.155%; on the other hand, Liaoning has been affected by the supply-side structural reform due to the high number of heavy industrial enterprises, resulting in slow growth in the level of resilience. Therefore, regional governments should strengthen top-level design to promote the healthy development of regional digital innovation ecosystems and take into account their development situations.

5.1.2. Analysis of the Spatial Evolution of the Resilience Level

Table 3. The results of global Moran’s I from 2011 to 2021.

Year	2011	2012	2013	2014	2015	2016	2017	2018	2019	2020	2021
Moran’s I	0.228 **	0.251 ***	0.243 **	0.223 **	0.208 **	0.191 **	0.162 **	0.124 **	0.108 *	0.096 *	0.128 *
p-value	0.019	0.009	0.014	0.016	0.024	0.029	0.039	0.046	0.087	0.095	0.072

Note: ***, **, and * denote significance at the 1%, 5%, and 10% levels, respectively.

and Figure 3 show the calculated global Moran’s I values for the resilience levels. The global Moran’s I was greater than zero, decreasing by 43.860%—from 0.228 (2011) to 0.128 (2021)—which indicates that there is a positive correlation between resilience levels and the relationship is gradually weakening, which corresponds to the inter-regional “low coupling” characteristic of the resilience level. The spatial distribution pattern of resilience level shifts from “low level and low gap” to “high level and high gap,” and the polarization phenomenon among regional digital innovation ecosystems is prominent. At the same time, the global Moran’s I reached a minimum value of 0.096 in 2020 and began to increase in 2021, indicating that the polarization effect of resilience levels has improved and shows a positive trend.

We calculated the local Moran’s I to summarize the clustering pattern of the resilience levels; the results are presented in

Table 4. Provinces and cities included in each quadrant in 2011, 2016, and 2021.

Year	H–H (The First Quadrant)	L–H (The Second Quadrant)	L–L (The Third Quadrant)	H–L (The Fourth Quadrant)
2011	Beijing, Hebei, Shanghai, Jiangsu, Zhejiang, Shandong, Henan	Tianjin, Anhui, Fujian, Jiangxi, Hunan, Guangxi, Hainan, Chongqing	Shanxi, Inner Mongolia, Jilin, Heilongjiang, Guizhou, Yunnan, Shaanxi, Gansu, Qinghai, Ningxia, Xinjiang	Liaoning, Hubei, Guangdong, Sichuan
2016	Beijing, Tianjin, Shanghai, Jiangsu, Zhejiang, Shandong, Henan	Hebei, Anhui, Fujian, Jiangxi, Hunan, Guangxi, Hainan	Shanxi, Inner Mongolia, Liaoning, Jilin, Heilongjiang, Chongqing, Guizhou, Yunnan, Shaanxi, Gansu, Qinghai, Ningxia, Xinjiang	Hubei, Guangdong, Sichuan
2021	Shanghai, Jiangsu, Zhejiang, Anhui, Fujian, Shandong, Henan, Hubei, Hunan	Tianjin, Hebei, Jiangxi, Guangxi, Hainan, Chongqing	Shanxi, Inner Mongolia, Liaoning, Jilin, Heilongjiang, Guizhou, Yunnan, Gansu, Qinghai, Ningxia, Xinjiang	Beijing, Guangdong, Sichuan, Shaanxi

. The “high–high” quadrant expanded during the survey period, and the distribution range was mainly concentrated on the East Coast and the middle reaches of the Yangtze River. The “low–high” quadrant decreased in scope, and the distribution range was more fragmented. The “low–low” quadrant was mainly distributed in the Northeast and the Great Northwest, where resilience levels were low. The “high–low” quadrant mainly included provinces and cities with higher resilience levels than surrounding regions such as Guangdong and Sichuan. The expansion of the scope of the “high–high” quadrant indicates that the resilience level of some provinces and cities has grown rapidly and that the overall resilience level has improved. For example, Anhui and Hubei are in the Yangtze River Delta economic zone, driven by Jiangsu, Zhejiang, and other developed provinces, the regional innovation level has improved rapidly and thus entered the “high–high” quadrant. Fujian and Hunan have also successfully entered the “high–high” quadrant from the “low–high” quadrant, driven by the Pearl River Delta region.

5.2. Influencing Factors Analysis for the Resilience of Regional Digital Innovation Ecosystems

5.2.1. Obstacle Factors Analysis

Figure 4 shows the obstacle levels of the indicators for each stage of the digital innovation ecosystem in the 30 regions for 2011, 2016, and 2021. In 2021, the main obstacles to resilience development in the current regional digital innovation ecosystem are the state and response indicators, whereas pressure indicators have a relatively low impact on resilience, accounting for 3.420%, 38.402%, and 58.178% of Stages 1, 2, and 3, respectively. Guangdong had the highest proportion of pressure barriers, which is relatively abnormal, indicating that the digital innovation ecosystem in the region needs to prevent potential risks and ensure that the system operates smoothly. Beijing had a low level of state barriers and the highest proportion of response barriers, indicating that although its system has a large scale and strong stability, its response and recovery capacities to external shocks are insufficient. From the perspective of change, the proportion of pressure disorders increased from 3.180% (2011) to 3.420% (2021) during the survey period, indicating that the instability of the external environment in the past decade hindered improvements in resilience to some extent. The distribution of obstacles in most provinces and cities was relatively stable, and the percentage of obstacles changed slightly at each stage. Guangdong was abnormal: the pressure and state obstacle degrees increased and decreased significantly, respectively; thus, with improvements in the system development level, the risk of external shocks in the region also increased. Accordingly, early warning emergency systems need to be strengthened in the future, along with relevant actions taken in advance, to ensure the stable development of the regional digital innovation ecosystem.

Specifically, indicators with obstacle degrees greater than 5% were analyzed as obstacle factors, as shown in Figure 5. The results show that there are seven obstacles, all of which belong to the state and response stages, indicating that the pressure indicators have a relatively small hindering effect on resilience. Among them, the indicator “S4: Technology contract turnover” was always the most significant obstacle factor, and the obstacle degree exceeded 15% at all time points, seriously hindering the improvement in the resilience level of the regional digital innovation ecosystem. Therefore, provinces and cities should focus on the transformation and application of innovation achievements and further improve the resilience level of the regional digital innovation ecosystem by reducing the barrier degree of the obstacle indicators. From the perspective of change, the change in obstacle degree of each indicator was relatively small, indicating that the hindering effect of obstacle factors on resilience has not been improved; therefore, provinces and cities should apply precise measures and propose suitable solutions for the above obstacle factors to promote the healthy and sustainable development of regional digital innovation ecosystems.

5.2.2. Driving Factors Analysis

A regional digital innovation ecosystem is a multifaceted system that includes various influential factors. Drawing on relevant research results [50,51], we selected industrial structure, regional urbanization, human capital, regional innovation, digital industrialization, and industry digitization as explanatory variables to investigate the factors that significantly impact resilience level, as shown in

Table 5. Resilience level dynamic factor index system.

	Variable Name	Variable Description	Variable Symbol	Unit
Explained variable	Regional digital innovation ecosystem resilience	Describes the resilience of the system to external shocks	RES	-
Explanatory variables	Industrial structure	Indicates the composition of various industries	Is	-
	Regional urbanization	Describes the level of urbanization construction	Ru	-
	Human capital	Expressed as a proportion of the population with a graduate degree	Hc	%
	Regional innovation	R&D investment intensity	Ri	%
	Digital industrialization	Expressed as total telecom business	Di	100 million CNY
	Industry digitization	The proportion of digital industry enterprises in enterprises above the designated size	Id	%

. Among these variables, industrial structure denotes the interconnectivity and proportion of different industries, with enterprises as the principal innovators playing a crucial role in the regional digital innovation ecosystem. Regional urbanization reflects the extent of urban development and indicates a region’s economic progress. Human capital highlights the human resources a region possesses and is an essential indicator of knowledge output. Additionally, regional innovation denotes the level of active innovation activities that occur and is significant for system resilience. The digital economy is intricately interlinked with the regional digital innovation ecosystem and enhancing digital innovation contributes to an improved level of resilience. Consequently, digital industrialization and industry digitization variables are employed to illustrate the level of digital economic development. All original data on the explanatory variables were sourced from the publicly accessible statistical yearbook published by the National Bureau of Statistics of China, as shown in

Table 6. Results of descriptive statistics of variables.

Variable Symbol	Sample Size	Maximum Value	Minimum Value	Average Value	Standard Deviation	Median
RES	330	0.750	0.073	0.197	0.100	0.169
Is	330	361,307.088	53,288.142	126,279.977	52,150.427	112,595.401
Ru	330	0.896	0.344	0.596	0.121	0.581
Hc	330	9.304	0.064	0.661	1.095	0.401
Ri	330	6.530	0.410	1.758	1.139	1.435
Di	330	19,323.700	45.890	1825.594	2513.451	753.840
Id	330	14.200	0.010	2.851	2.677	1.982

.

To analyze the role of the explanatory variables in influencing the resilience of the regional digital innovation ecosystem, we established a measurement model for the driving factors of regional digital innovation ecosystem resilience based on panel data from 30 provinces and cities in China from 2011 to 2021. In the formula, $LnRES$ is the explained variable; $LnIs$, $LnRu$, $LnHc$, $LnRi$, $LnDi$, and $LnId$ are the explanatory variables; $\delta$ is the individual fixed effect; and $\theta$ is a random interference term. We lagged the explanatory variables by one year to avoid endogeneity. As shown in Table 6, the minimum values of all variables were greater than zero; therefore, all variables were calculated by taking logarithms to reduce the effect of the heteroskedasticity problem on the regression. Considering that there may be multicollinearity problems between the variables which reduce the reliability of the estimation results, we calculated the correlation coefficients between the variables and created a correlation matrix. As shown in

Table 7. Correlation matrix.

Variable	RES	Is	Ru	Hc	Ri	Di	Id
RES	1.000 ***	0.448 ***	0.452 ***	0.200 ***	0.542 ***	0.629 ***	0.507 ***
Is	0.448 ***	1.000 ***	0.590 ***	0.376 ***	0.405 ***	0.368 ***	0.329 ***
Ru	0.452 ***	0.590 ***	1.000 ***	0.463 ***	0.520 ***	0.281 ***	0.364 ***
Hc	0.200 ***	0.376 ***	0.463 ***	1.000 ***	0.452 ***	0.083 **	0.203 ***
Ri	0.542 ***	0.405 ***	0.520 ***	0.452 ***	1.000 ***	0.309 ***	0.544 ***
Di	0.629 ***	0.368 ***	0.281 ***	0.083 **	0.309 ***	1.000 ***	0.312 ***
Id	0.507 ***	0.329 ***	0.364 ***	0.203 ***	0.544 ***	0.312 ***	1.000 ***

Note: *** and ** denote significance at the 1% and 5% levels.

, the absolute values of the correlation coefficients between most of the variables were less than 0.600, indicating that there was no significant multicollinearity between the variables.

$$Ln{RES}_t={\alpha }_0+{\alpha }_{1}Ln{Is}_{t-1}+{\alpha }_{2}Ln{Ru}_{t-1}+{\alpha }_{3}Ln{Hc}_{t-1}+{\alpha }_{4}Ln{Ri}_{t-1}+{\alpha }_{5}Ln{Di}_{t-1}+{\alpha }_{6}Ln{Id}_{t-1}+\delta +{\theta }_t$$

(20)

Table 8. Regression results.

	Full Sample	Northern Region	Southern Region
	Fixed Effect Model	Fixed Effect Model	Fixed Effect Model	Random Effect Model
LnIs	0.266 *** (5.330)	0.127 * (1.903)	0.528 *** (7.185)	0.442 *** (5.664)
LnRu	1.671 *** (10.335)	2.265 *** (10.419)	0.606 ** (2.570)	0.250 (1.012)
LnHc	0.038 ** (2.413)	0.058 ** (2.524)	−0.001 (−0.038)	−0.033 * (−1.841)
LnRi	0.038 (0.738)	−0.144 ** (−2.082)	0.208 ** (2.469)	0.212 *** (3.773)
LnDi	0.050 *** (3.635)	0.033 (1.648)	0.051 *** (2.850)	0.097 *** (5.791)
LnId	−0.005 (−0.396)	−0.009 (−0.568)	0.030 (1.366)	0.034 * (1.779)
_cons	−4.329 *** (−7.675)	−2.394 *** (−3.331)	−7.979 *** (−9.057)	−7.535 *** (−7.831)
N	300	130	170	170
F	323.917	127.016	254.429	223.609
P	0.000	0.000	0.000	0.000
R2	0.880	0.873	0.912	0.907
Hausman test	p = 0.000	p = 0.000	p = 1.000

Note: ***, **, and * denote significance at the 1%, 5%, and 10% levels, respectively; T-values are shown in parentheses.

shows that the results of the Hausman test were significant, thus prompting the selection of a fixed-effects model for analysis. After treating the independent variables with a one-year lag, most variables were significant, indicating that the results were robust. Based on the regression outcomes of the fixed effects, industrial structure, regional urbanization, human capital, and digital industrialization had a significantly positive impact on the explained variable, indicating that these variables act as drivers of regional digital innovation ecosystem resilience. The regression coefficient of LnRu is the highest, indicating that regional urbanization makes the strongest contribution to the resilient development of the regional digital innovation ecosystem. Therefore, provinces and cities should actively and steadily promote urbanization, closely integrate regional economic development and industrial layouts, and concentrate on improving the quality of urbanization. We partitioned China’s geographical area into two parts to conduct regional heterogeneity testing. The northern region included 13 provinces and cities: Beijing, Tianjin, Hebei, Shanxi, Inner Mongolia, Liaoning, Jilin, Heilongjiang, Shaanxi, Gansu, Qinghai, Ningxia, and Xinjiang. The remaining 17 provinces and cities belonged to the southern region. As can be seen from Table 8, the regression in the northern region employs a fixed effects model, and the findings reveal that the regional digital innovation ecosystem resilience is significantly and positively impacted by industrial structure, regional urbanization, human capital, and regional innovation. The coefficient of LnRu in this region is higher than that in the overall sample, indicating that the positive effect of urbanization on resilience is more pronounced in the northern region than in the entire sample. In the southern region, a random-effects model was selected for analysis, based on the results of the Hausman test. The results show that the regression coefficients of lnIs, lnHc, lnRi, lnDi, and lnId were significant. Taken together, these results indicate that there are differences between the northern and southern regions. For example, human capital shows a positive contribution in the northern region and a negative effect in the southern region, which may be related to the different resource conditions and developmental statuses of the two regions. Simultaneously, there are similarities between the two regions. For example, industrial structure is significant in both regions, indicating that the positive contribution of industrial structure to regional digital innovation ecosystem resilience is universal in China and that regional governments should accelerate the transformation of economic development, promote the optimization and upgrading of industrial structure, and lay a solid foundation for the regional digital innovation ecosystem to enhance resilience.

6. Research Conclusions, Suggestions, and Research Limitations

6.1. Research Conclusions

This study conducted an empirical analysis based on panel data from 30 provinces and cities in China from 2011 to 2021. Specifically, we explained the concept of regional digital innovation ecosystem resilience, built an evaluation index system based on it, measured the resilience level of regional digital innovation ecosystems using the global entropy method and TOPSIS model, and analyzed the results with Moran’s I. Additionally, we used the obstacle and panel models to identify and discuss the obstacles and driving factors affecting the resilience level, respectively.

The research indicates:

• During the survey period, China’s regional digital innovation ecosystem resilience level was improving steadily, with an increase of 128.571%. It also exhibited a spatial distribution pattern of “coastal high–central flat–northwest depression.” The resilience level of intra-regional “high cohesion” and inter-regional “low coupling” characteristics were prominent, and inter-regional differences were significant. Therefore, this hypothesis was verified.
• Global Moran’s I was always positive and showed a decreasing trend during the survey period, and the correlation between regions gradually weakened. The spatial pattern changed from “low level, low gap” to “high level, high gap,” and most of the provinces and cities were located in the “high–high” and “low–low” quadrants. The scope of the “high–high” quadrant and the “low–high” quadrant has been expanding and narrowing, respectively.
• State and response indicators are the main obstacles affecting the resilience of the regional digital innovation ecosystem, with a relatively small proportion of pressure indicator obstacles. During the survey period, the changes in obstacle levels in various provinces and cities were relatively small, and the obstacle levels of the pressure indicators increased. Indicators such as “technology contract turnover” and “total number of patents granted” had a high obstacle degree and were the main obstacle factors for the resilient development of the regional digital innovation ecosystem.
• Overall, industrial structure, regional urbanization, human capital, and digital industrialization make significant positive contributions to system resilience. Through regional heterogeneity analysis, we found that industrial structure, regional urbanization, human capital, and regional innovation factors were significant in the northern region, whereas in the southern region, industrial structure, regional innovation, digital industrialization, and industry digitization had positive driving effects on the resilience level of regional digital innovation ecosystems.

6.2. Suggestions and Advice

Based on these results, we propose the following countermeasures and suggestions for improving the resilience level of the regional digital innovation ecosystem and reducing the difference in resilience levels between regions.

• Overall, the top-level design should be strengthened to seek a path for resilient development at the highest level. Specifically, regional integration construction strategies should be actively pursued to improve support for the western and northeastern regions and promote the common development of regional digital innovation ecosystems, which bring about a series of governance issues related to digital technology and data [52]. Specifically, it is necessary to strengthen the government’s guiding and supervisory responsibilities, combine endowment conditions with the external environment, formulate resilience enhancement according to local condition policies, and simultaneously develop a resilience assessment flexibility space to avoid blind and disorderly upgrades and promote the healthy and sustainable development of the digital innovation industry and regional digital innovation ecosystem.
• Eastern regions, such as Jiangsu, Shanghai, and Guangdong, should take advantage of their resources, focus on the construction of high-level innovation platforms, promote the development of high-tech industries with core intellectual property rights, build digital industrial parks (e.g., electronic information and high-tech manufacturing), and improve and sound the innovation industry chain to build an industrial system with digital innovation as the core. It is also necessary to promote cross-regional digital innovation transactions and facilitate the flow and sharing of innovation results to drive the development of neighboring regions.
• For the central regions (e.g., Jiangxi, Hunan, and Henan), the role of articulation should be given full play. On one hand, they should effectively undertake the digital innovation industry in high-resilience regions, pay attention to the cultivation and introduction of innovative talent, create a soft environment for talent development, and establish a talent base for the development of regional digital innovation ecosystems. On the other hand, they should strengthen innovation interaction with the western and northeastern regions, improve the flow mechanism of innovation resources, and promote inter-regional collaborative innovation and value co-creation.
• The western and northeastern regions (e.g., Xinjiang, Qinhai, and Liaoning), where resilience levels are low, should coordinate innovation resources and optimize the allocation of innovation factors. They should also build an “Internet+” innovation exchange platform by relying on information technologies such as “cloud computing” and “blockchain,” strengthen the innovation interaction with neighboring regions, and promote the coordinated development of regional innovation. Simultaneously, they should build a multilevel and multichannel innovation financing system using interest rate subsidies, guarantees, and tax breaks to support the development of independent regional enterprises and regional innovation.
• Provinces and cities should pay attention to the various factors that affect regional digital innovation ecosystem resilience. On the one hand, for obstacle factors, provinces and cities should trace the root of barriers and reduce the obstacle degree of indicators fundamentally. For example, as the obstacle degree of the technology contract turnover indicator is always high, the government should promote the transformation and landing of innovation results; accelerate the construction of regional innovation result trading platforms; extend the innovation value chain; enhance the synergistic relationship among innovation subjects such as industry, university, and research institutes; and provide conditions for the market-oriented development of innovation results. On the other hand, regions should guide the development of the regional digital innovation ecosystem around driving factors such as digital industrialization and regional urbanization. For example, provinces and cities should vigorously develop the digital economy to support digital innovation activities and focus on regional urbanization to promote the positive role of driving factors in improving the resilience level of the regional digital innovation ecosystem.

6.3. Research Limitations

Although this study analyses and explains the concept of regional digital innovation ecosystem resilience, additional explanations can be derived from actual cases in the future to make the research more realistic. Additionally, the resilience evaluation index system can be improved further, and relevant and reasonable indicators can be added in the future. Finally, based on the analysis of influencing factors, the deep-seated mechanism of action can be further studied to enhance the significance of this research.