The Spatiotemporal Patterns and Driving Mechanism of the Synergistic Effects in Industrial Green Production

Abstract

Making full utilisation of the synergies that exist among the various stages of industrial green production is beneficial to the realisation of the dual-carbon goal. However, the synergistic effects among the three stages of industrial green production have not yet been explored in depth from a microscopic perspective. Based on the analytic hierarchy process, the entropy weighting method, the coupled synergy degree model, the spatial autocorrelation test, and the geographically weighted regression model (GWR), the spatiotemporal evolution characteristics and driving mechanism of the synergistic effects among the three stages of industrial green production were explored by utilising the relevant data of industrial enterprises in 30 provinces of China from 2012 to 2022. The results showed that the synergistic effect of industrial green production exhibited an upward trend over time, and displayed a regional distribution characteristic of decreasing from east to west. The spatial differences in the synergistic effects of industrial green production gradually narrowed and the number of provinces with high–high (H-H) agglomerations increased. The level of digital economy development, the urbanisation level, the optimisation of industrial structure, the level of green credit, and the intensity of environmental regulation were the main driving factors for the synergistic effects of industrial green production, and there were significant spatial differences. This study provides a basis for the formulation of differentiated regional green development policies from the perspective of synergizing the various stages of industrial green production.

1. Introduction

China’s industry is responsible for approximately 70% of the country’s carbon emissions and more than 40% of its major pollutant emissions (e.g., SO₂ and NOx). According to the latest data from the National Bureau of Statistics in 2022, industrial output accounted for 33.2% of GDP, but its energy consumption accounted for approximately 65% of the country’s total energy consumption. Of the industrial technological reforms in 2024, less than 20% of the investment was for energy-saving and environmental upgrades, and the rest was for capacity expansion. The rough development model puts severe pressure on China’s environment [1,2]. Addressing carbon emissions from industry is crucial for China to meet the goal of achieving carbon peaking and carbon neutrality. Industrial green production is a critical path to reducing carbon emissions in the industrial sector by realising low-carbon development through technological innovation and management optimisation. Therefore, improving the industrial green production capacity has become a growing focus of attention.

At present, the space for environmental improvement in the traditional terminal treatment modes is increasingly restricted and this predicament is driving the transformation of industrial green production from passive pollution control to whole-process management and control [3,4]. However, existing research on industrial green production has emphasised external characterisation, for example, industrial green development [5,6,7], industrial green transformation [8,9], industrial low-carbon development [10,11,12], and green manufacturing [13]. The formation elements in industrial green production and the interaction process among the production stages of the whole life cycle have not yet been elucidated, which means there is great potential for improving the capacity of industrial green production.

From the perspective of the whole life cycle, industrial green production includes three stages: the source green production management stage (SGPMS), the process green production optimisation stage (PGPOS), and the end green production control stage (EGPCS) [13,14]. Industrial green production forms a deeply coupled, synergistic network through the interaction of material, energy, and information across production stages. Specifically, SGPMS sets environmental benchmarks for PGPOS and EGPCS through eco-design, clean material substitution, and green process selection [15]. Through energy consumption control, circular economy practice, and process innovation, PGPOS not only optimises the management decisions of SGPMS, such as feedstock selection and process parameter setting, but also improves the management efficiency of EGPCS, such as waste toxicity reduction and treatment cost reduction [16]. EGPCS builds negative feedback mechanisms through waste recycling, carbon capture, and advanced pollution control to effectively offset the residual environmental impacts of SGPMS and PGPOS. Meanwhile, EGPCS optimises the production process design of SGPMS and PGPOS through reverse analysis of material flow data [17,18].

Under the strategic guidance of ecological civilisation construction and the targets for carbon peaking and carbon neutrality, China’s attention to the synergy of industrial green production has increased greatly. The Implementation Programme for Green and Low Carbon Standardisation Work (2025), the Announcement on Zero Carbon Industrial Park Construction (2025), and the ‘14th Five-Year Plan’ (2021–2025) for the Green Development of Industry explicitly stated that it was necessary to establish a synergistic governance framework for the ‘entire industrial chain’ to promote the green transformation of industry. Industrial green production synergy among the three stages reduces resource consumption and pollution emissions of the production system through technological innovation and optimal resource allocation [17]. It promotes the transformation of the traditional industrial production model to an environmentally friendly one through cross-sectoral technology integration, cross-industry chain technology integration, and institutional innovation. The synergy among the three stages of industrial green production has become the key to breaking through environmental constraints [3,10]. Exploring the evolutionary characteristics and driving mechanism of the synergistic effects in industrial green production not only facilitates the improvement of the synergistic efficiency of industrial green production but also has profound significance for the ecologisation of regional industries.

The synergistic effects of industrial macro-development elements have attracted a large number of studies, focusing, for example, on economy and green development [19,20], technological innovation and green development [21,22], finance and green development [12,23], digitisation and green development, and policy synergistic effects [11,24]. Research on synergistic effects in industrial green production has focused on elements in a single production stage, such as the synergistic effects of pollution and carbon reduction [9,25,26], and synergistic innovation effects [27,28]. The methodologies used in the above studies include the entropy weighing method, the principal component analysis method, and the coupled synergy model. However, there are fewer studies on the synergistic effects among the three stages of industrial green production from a microscopic perspective. As a result, there is a lack of comprehensive quantification of the synergies among industrial production stages, making it difficult to reveal synergistic mechanisms across the whole life cycle of industry.

The existing studies have mainly analysed the driving factors of synergistic effects for industrial macro-development elements. The multifactor analysis methods mainly include the ordinary least squares (OLS) model, censored regression (Tobit) model, stochastic impacts by regression on population, affluence, and technology (STIRPAT) model, geographically weighted regression (GWR) model, and geographically and temporally weighted regression (GTWR) model [29,30,31]. Environmental driving factors have focused on physical geographical dimensions and socioeconomic dimensions, such as hydrometeorological variables, geomorphological features, innovation, urbanisation process, and industrial structure [32,33,34,35]. However, the driving factors of green production synergistic effects at the industrial micro level are yet to be explored.

Synergistic effects of industrial green production are influenced by a combination of basic support conditions, institutional environment, and market environment [36,37,38]. Key driving factors in this process include digital economy development, environmental regulatory intensity, green credit, urbanisation, and industrial structure optimisation. Digital economy development significantly enhances the synergistic effects of industrial green production by the application of advanced technologies. For example, Vadurin et al. [39] argued that digital economic development precisely controlled pollution emissions through the Internet. Anser et al. [40] found that industrial energy efficiency was driven through smart manufacturing. Environmental regulatory intensity reconfigures industrial enterprises’ production decisions through the constraining mechanism of increasing the cost of non-compliance [41] and the inducing mechanism of creating a green premium [42], which in turn facilitates the synergistic effects of industrial green production. Green credit directs the allocation of industrially produced capital to environmentally friendly projects [43]. Urbanisation facilitates the sharing of green infrastructure and the diffusion of green technologies, which in turn strengthen the synergistic effects of industrial green production. Industrial structure optimisation promotes resource allocation efficiency across sectors of industrial production [44]. However, the spatial heterogeneity of how these driving factors influence the synergistic effects of industrial green production requires further empirical verification.

In summary, research on industrial green production and industrial synergistic effects has achieved rich results, but there are still deficiencies yet to be investigated. First, while the external characterisation of industrial green production has been extensively explored, they have neglected its basic components and the complex interactions among the stages of the whole production life cycle. Second, existing evaluations of industrial synergistic effects have mainly focused on industrial development elements at the macro level, while research on micro-level synergistic effects among the three stages of industrial green production remains notably scarce. Finally, there are large regional differences in the synergistic development of industrial green production across China’s provinces, but the causes of these spatial differences are not yet clear. The research gaps constrain the scientific formulation of differentiated regional policies.

Given the above research limitations, we addressed the spatial pattern and driving factors of the synergistic effects among the three stages of industrial green production. The possible marginal contributions of this study were threefold. First, based on the characteristics of the whole life cycle of industrial production, an indicator system of industry green production was constructed. The synergistic effects among the three stages of industrial green production were measured using the coupled synergy model. The spatial and temporal evolution characteristics of the synergistic effects in industrial green production were analysed to provide theoretical insights for improving industrial green development. Second, the changes in the spatial pattern of the synergistic effects of industrial green production were analysed via Moran’s I index and the spatial geographic standard deviation ellipse to provide guidance for optimising the synergistic path of industrial green production. Third, the driving factors of the synergistic effects in industrial green production and their influence on spatial heterogeneity were investigated via the GWR model, thereby providing decision-making references for the formulation of a synergistic policy of industrial green production in accordance with local conditions.

The rest of this study is organised as follows: the Section 2 presents the indicator system and research methodologies, and the Section 3 presents the empirical analysis. The Section 4 presents the discussion. The Section 5 presents the research conclusions and implications.

2. Indicator System and Research Methodologies

2.1. Industrial Green Production Indicator System

Given that the industrial green production system involves production management, resource efficiency, environmental impact, and economic benefits, it cannot be accurately described by a single quantitative indicator [13,45]. Therefore, in the construction of the indicator system, references were made to the results of the Sustainable Development Reporting Guidelines (G4) issued by the Global Reporting Initiative, the Green Development Indicator System established by Chinese government departments, and the Industrial Green Transformation Index compiled by Chinese government departments. In addition, following the principles of objectivity, comprehensiveness, and accessibility, while drawing on the indicator setting methodology of Yang et al. and Ijaz et al. [13,45], an industrial green production indicator system was constructed, as shown in

Table 1. Indicator system for China’s industrial green production.

Level 1 Indicators	Level 2 Indicators	Interpretation of Indicators	Property	Entropy Weights	AHP Weights	Comprehensive Weights
SGPMS	Green production support	Number of industrial green policies (units)	+	0.169	0.517	0.343
		Total investment in urban environmental infrastructure construction (billion yuan)	+	0.103	0.306	0.205
	Industrial development capacity	Expenditure on new product development in industrial enterprises (10 thousand yuan)	+	0.298	0.101	0.199
		Expenditure on technology introduction in industrial enterprises (10 thousand yuan)	+	0.429	0.077	0.253
PGPOS	Utilisation of energy and waste	Comprehensive energy consumption per unit of industrial GDP (10 thousand tons of standard coal/billion yuan)	−	0.118	0.171	0.144
		Coal consumption per unit of industrial GDP (10 thousand tons/billion yuan)	−	0.074	0.168	0.121
		Clean new energy generation per unit of industrial GDP (100 million kilowatt hours/billion yuan)	+	0.007	0.184	0.096
		Industrial water volume conserved (10 thousand cubic metres)	+	0.188	0.064	0.126
		Industrial water volume reused (10 thousand cubic metres)	+	0.153	0.063	0.108
		Integrated utilisation of general industrial solid waste (10 thousand tons)	+	0.065	0.057	0.061
	Innovation capacity	Full-time equivalent of R&D personnel in industrial enterprises (person-years)	+	0.155	0.105	0.130
		Technological transformation expenditure of industrial enterprises (10 thousand yuan)	+	0.094	0.086	0.090
		Industrial enterprises’ R&D expenditures (10 thousand yuan)	+	0.146	0.102	0.124
EGPCS	Industrial growth quality	Growth rate of industrial added value (%)	+	0.151	0.173	0.162
		Sales revenue of new products per unit of industrial added value (10 thousand yuan)	+	0.236	0.210	0.223
	Industrial pollution control and investment	Wastewater emissions from industry (10 thousand tons)	−	0.017	0.052	0.034
		SO2 emissions from industry (10 thousand tons)	−	0.016	0.052	0.034
		Fumes emissions from industry (10 thousand tons)	−	0.010	0.047	0.028
		General solid waste emissions from industry (10 thousand tons)	−	0.014	0.042	0.028
		Industrial pollution control investment (10 thousand yuan)	+	0.125	0.076	0.100
		Completed annual investment in industrial waste gas treatment (10 thousand yuan)	+	0.157	0.088	0.123
	Carbon sink capacity	Forest coverage rate (%)	+	0.131	0.137	0.134
		Greening coverage rate of built-up areas (%)	+	0.144	0.123	0.134

.

According to the concept of the whole life cycle management, industrial green production can better achieve a balance between environmental and economic benefits through the synergistic effects among SGPMS, PGPOS, and EGPCS. Therefore, an industrial green production indicator system of the whole life cycle, including SGPMS, PGPOS, and EGPCS, was constructed. Compared with existing studies, this study combined the dual needs of industrial structural transformation and green development, and added the indicators of SGPMS, i.e., industrial development capacity and green production support. The indicators of EGPCS were extended, i.e., carbon sink capacity. This can more comprehensively reflect the synergistic management effectiveness of the whole life cycle in industrial green production, characterised by source management–process optimisation–end control. The logical interpretation of the industrial green production indicator system was as follows:

(1)

The core components in SGPMS were green production support and industrial development capacity. Green production support was characterised mainly by the amount of green policy and the total investment in urban environmental infrastructure, which were derived from the Peking University Legal Information Database and the China Urban Construction Statistical Yearbook, respectively. SGPMS was influenced by green production support through the dual path of ‘institutional incentives and infrastructural safeguards’ [45,46]. The key indicators of industrial development capacity were expenditures on new product development and expenditures on technology introduction, which were obtained from the China Industrial Statistical Yearbook. The mechanism of industrial development capacity in SGPMS was mainly reflected in the reduction in resource consumption intensity and pollution emissions through the application of green technologies [19,47].

(2)

The core components in PGPOS were the utilisation of energy and waste, as well as innovation capacity. The data on the utilisation of energy and waste were obtained from the China Energy Statistical Yearbook, the China Electricity Statistical Yearbook, provincial statistical yearbooks, and the China Industrial Statistical Yearbook. Indicators of energy utilisation determined the pollution emissions from the industrial production process through the intensity of energy consumption, the optimisation of the energy structure, and the conservation and recycling of energy [48,49]. The data for innovation capacity were from the China Science and Technology Statistical Yearbook and the China Statistical Yearbook. Studies have shown that technician reserves and investment in R&D have a significant positive impact on the green growth of PGPOS [7,47].

(3)

The core components in EGPCS were industrial growth quality, industrial pollution control and investment, and carbon sink capacity. In this study, industrial growth quality was represented by the growth rate of industrial added value and the sales revenue of new products per unit of industrial added value, with data from the China Industrial Statistical Yearbook. The improvement in output efficiency presented by the indicators of industrial growth quality reflected the enhancement of capacity in EGPCS [50]. Industrial pollution control and investment were described by six indicators related to pollutant emissions and environmental investment, with data from the China Environmental Statistics Yearbook. Research has shown that industrial pollution control and investment have a decisive effect on pollution reduction in EGPCS [5,51]. The core evaluation indicators of carbon sinks were the forest coverage rate and the greening coverage rate of built-up areas, with data from the China Urban Construction Statistical Yearbook, the China Statistical Yearbook, and provincial statistical yearbooks. The forest coverage rate was the core carrier of natural carbon sinks, while the greening coverage rate of built-up areas reflected the construction level of artificial carbon sinks. Their ecological regulation function significantly reduced pollution emissions in EGPCS [22,46].

2.2. Methodologies

2.2.1. Indicator Weighting Methods

Considering the high subjectivity and arbitrariness of the subjective weighting method, as well as the mismatch between the weighted results and the theoretical importance of the objective weighting method, it was necessary to use a combination of subjective and objective assignment methods [52]. Therefore, industrial green production capacity in SGPMS, PGPOS, and EGPCS was measured using an integrated method that combined the analytic hierarchy process (AHP) with the entropy weighting method.

2.2.2. Synergetic Model

The measurement of synergistic effects usually adopts a system dynamics model, a super-efficient SBM model, and a coupled synergy model [3,30,31]. Compared with the previous two models, the coupled synergy model can accurately quantify the dynamic equilibrium of system interactions and effectively reveal the synergistic evolution mechanism of multi-dimensional elements [53]. Therefore, the coupled synergy degree calculated by a synergetic model was used in this study to explore the synergistic effects among the three stages in industrial green production. First, the coupling degree and comprehensive development index among the subsystems were calculated with the following equations:

$$C_1={\left[{\displaystyle \frac{F_1\times F_2}{{\left(\frac{F_1+F_2}{2}\right)}^2}}\right]}^{{\displaystyle \frac{1}{2}}};C_2={\left[{\displaystyle \frac{F_1\times F_3}{{\left(\frac{F_1+F_3}{2}\right)}^2}}\right]}^{{\displaystyle \frac{1}{2}}};C_3={\left[{\displaystyle \frac{F_2\times F_3}{{\left(\frac{F_2+F_3}{2}\right)}^2}}\right]}^{{\displaystyle \frac{1}{2}}};C_4={\left[{\displaystyle \frac{F_1\times F_2\times F_3}{{\left(\frac{F_1+F_2+F_3}{3}\right)}^3}}\right]}^{{\displaystyle \frac{1}{3}}}$$

(1)

$$T_1={\beta }_1{F}_1+{\beta }_2{F}_2;T_2={\delta }_1{F}_1+{\delta }_2{F}_3;T_3={\gamma }_1{F}_2+{\gamma }_2{F}_3;T_4={\eta }_1{F}_1+{\eta }_2{F}_2+{\eta }_3{F}_3$$

(2)

where $F_1$, $F_2$, and $F_3$ represent the values for the SGPMS, PGPOS, and EGPCS subsystem levels. $C_1$, $C_2$, and $C_3$ represent the coupling degree between two subsystems of industrial green production. $C_4$ represents the coupling degree among the three subsystems of industrial green production. $T$ represents the comprehensive development index for the subsystem. $\beta$, $\delta$, $\gamma$, and $\eta$ represent the subsystem weights. In reference to the experts’ group’s opinions, the subsystem weights were assigned as ${\beta }_1=0.46$, ${\beta }_2=0.54$; ${\delta }_1=0.46$, ${\delta }_2=0.54$; ${\gamma }_1=0.5$, ${\gamma }_2=0.5$; and ${\eta }_1=0.3$, ${\eta }_2=0.35$, ${\eta }_3=0.35$.

The system’s coupled synergy degree was subsequently calculated with the following equation:

$$D=\sqrt{C\times T_i}.$$

(3)

where $i=1,2,3,4$. D represents the coupled synergy degree, taking the value of 0~1; the larger the value of D, the higher the synergistic effect among the subsystems. Referring to the real situation of the industrial production process, the synergistic effect was categorised into 10 levels [54]. The classification levels are shown in

Table 2. Classification of synergistic effects in industrial green production.

Interval	Synergistic Effects Type	Interval	Synergistic Effects Type
(0.0, 0.1]	Extremely dysfunctional recession type	(0.5, 0.6]	Barely synergistic development type
(0.1, 0.2]	Severe dysregulation recession type	(0.6, 0.7]	Primary synergistic development type
(0.2, 0.3]	Moderately dysfunctional recession type	(0.7, 0.8]	Intermediate synergistic development type
(0.3, 0.4]	Mildly dysfunctional recession type	(0.8, 0.9]	Good synergy development type
(0.4, 0.5]	Hardly dysfunctional recession type	(0.9, 1.0]	Excellent synergistic development type

.

2.2.3. Spatial Autocorrelation and Standard Deviation Ellipse

The global Moran index was advantageous in reflecting the distribution characteristics of clustered, discrete, and stochastic for the synergistic effects of industrial green production in the global region [55]. Therefore, the global Moran index was employed to reveal the degree of spatial correlation of synergistic effects among the subsystems of industrial green production. As the global Moran index did not adequately reflect the extent and location of interprovincial agglomeration [56], in this study, we employed the local Moran index to further explore whether there was a spatial aggregation of synergistic effects in industrial green production. The formulas for the global Moran index and local Moran index are as follows, respectively:

$$\mathrm{Moran}’\mathrm{s}I={\displaystyle {\sum }_{i=1}^n{\displaystyle {\sum }_{j\ne 1}^n{W}_{ij}\times (x_i-\overline{x})\times (x_j-\overline{x})}}/{\displaystyle {\sum }_{i=1}^n{\displaystyle {\sum }_{j\ne 1}^n{(x_i-\overline{x})}^2}}$$

(4)

$$I_i=[(x_i-\overline{x})/S^2]\times {\displaystyle {\sum }_{j\ne 1}^n{W}_{ij}}\times (x_i-x)$$

(5)

where $W_{ij}$ represents the spatial weight matrix. $x_i$ and $x_j$ represent the industrial green production capacity in province $i$ and province $j$, respectively. $\overline{x}$ represents the average value of green production capacity in the industry, and $S^2$ is the discrete variance of $x_j$.

Based on the standard deviation ellipse, not only were the direction and discrete characteristics of the distribution of the synergistic effects in industrial green production explored, but their dynamic variation trend was also analysed. The spatial distribution parameters of the synergistic effects in industrial green production included the centre of gravity, long axis, short axis, and azimuth. The calculation formula can be found in the study of Duman et al. [57].

2.2.4. GWR

Compared with global regression methods, such as the OLS model and panel regression model, the GWR model can better reveal the local correlation characteristics of variable relationships with geospatial changes [31]. In analysing the reasons for the spatial heterogeneity of synergistic effects in industrial green production, the GWR model showed better explanatory power due to its spatially variable coefficient characteristics. [58]. Therefore, the distribution characteristics of spatial differences in the influence of various driving factors on the synergistic effects in industrial green production were revealed by the GWR model. The calculation formula is as follows:

$$y_i={\beta }_0(u_i,v_i)+{\displaystyle {\sum }_{k=1}^p{\beta }_k(u_i,v_i)x_{ik}}+{\epsilon }_i$$

(6)

where $y_i$ represents the degree of the synergistic effects in the industrial green production of province $i$. ${\beta }_0(u_i,v_i)$ is the intercept term. $(u_i,v_i)$ is the spatial geographic coordinates of province $i$. ${\beta }_k(u_i,v_i)$ is the $k$-th regression coefficient of location $(u_i,v_i)$. $x_{ik}$ is the $k$-th driving factor in province $i$. ${\epsilon }_i$ is the random error.

2.3. Data Sources

The data were mainly from the China Science and Technology Statistical Yearbook, the China Environmental Statistical Yearbook, the China Energy Statistical Yearbook, the China Industrial Statistical Yearbook, the China Electric Power Statistical Yearbook, the China Urban and Rural Construction Statistical Yearbook, the China Statistical Yearbook, statistical yearbooks of each province, and the Peking University Legal Information Platform. According to Pang et al. [59], this study divided mainland China into three regions: the eastern region, the central region, and the western region. Missing values (≤1% of total data) were handled via linear interpolation. The descriptive statistics of the main research variables are shown in

Table 3. Descriptive statistics for research variables.

Category	Variable	Mean	P50	SD	Min	Max
industrial green production capacity	Green production capacity in EGPCS	0.143	0.103	0.155	0.001	0.906
	Green production capacity in PGPOS	0.189	0.158	0.110	0.046	0.632
	Green production capacity in SGPMS	0.241	0.193	0.144	0.105	0.810
Synergistic effects	synergistic effects in industrial green production	0.402	0.381	0.131	0.133	0.910
Driving factors	Green Innovation (Ginnov)	8.488	8.496	1.339	4.466	11.207
	Level of transport infrastructure (Tinfra)	11.714	11.986	0.852	9.437	12.913
	Level of digital economy development (DEDL)	0.112	0.070	0.111	0.017	0.711
	Labour level (Labour)	7.601	7.658	0.768	5.545	8.864
	Per capita GDP (PGDP)	10.908	10.859	0.445	9.849	12.155
	Environmental regulation intensity (ERI)	0.260	0.232	0.111	0.105	0.758
	Level of green credit (GCL)	0.521	0.507	0.154	0.192	0.906
	Level of social consumption (Sconsum)	0.389	0.394	0.059	0.180	0.504
	Intensity of foreign direct investment (FDI)	0.265	0.145	0.268	0.008	1.354
	Urbanisation level (UR)	0.607	0.593	0.117	0.363	0.896
	Optimisation of industrial structure (ISO)	1.384	1.221	0.751	0.611	5.283

.

3. Empirical Analysis

3.1. Evolutionary Characteristics of Industrial Green Production Capacity

The evolutionary trend of industrial green production capacity is shown in Figure 1. Overall, industrial green production capacity increased rapidly from 0.437 to 0.700 during the study period, with a cumulative increase of 59.89%. The changes in industrial green production capacity regarding the subsystems are shown in Figure 1a. The results demonstrate that the industrial green production capacity of each subsystem exhibited the characteristics of EGPCS > PGPOS > SGPMS. From the empirical data in 2022, the compositional percentage of industrial green production capacity showed that the EGPCS subsystem had the highest percentage at 40.62%, while the EGPCS subsystem had the lowest percentage at 26.19%. This phenomenon was closely related to the payback period and environmental regulation. Specifically, SGPMS had a short payback period and fast short-term returns, while EGPCS had the longest payback period and a high investment threshold for green transformation. Under the pressure of environmental regulation, industrial enterprises tended to prioritise the improvement of SGPMS [51]. Figure 1b shows the changes in industrial green production capacity by region. The results indicate that green production capacity in industry was highest in the eastern region and lowest in the western region, which was strongly related to the industrial structure and energy consumption structure [60,61]. Compared to the western region, the eastern region had a better industrial structure and energy consumption structure. The eastern region was dominated by the service industry, with the tertiary industry accounting for 52.3%, while the western region was dominated by heavy industry, with the secondary industry accounting for 43.7%. Coal consumption in the western region accounted for 72.4%, while in the eastern region, the corresponding value was 49.8%. In addition, the proportion of renewable energy consumption in the eastern region was 26% higher than that in the western region. Therefore, the eastern region was superior in industrial green production.

3.2. Evolutionary Characteristics of Synergistic Effects in Industrial Green Production

The synergistic effects in industrial green production are shown in Figure 2. Overall, an increasing trend of the synergistic effects in industrial green production was evident from 2012 to 2022. Rising from 0.376 to 0.483, the synergistic effects crossed from the mildly dysfunctional recession type to the hardly dysfunctional recession type (orange curve). From a temporal perspective, although the synergistic effects in the regions of the east, the centre, and the west all showed a significant upwards trend, the synergistic effects in most of the regions were still of a hardly dysfunctional recession type in 2022. This phenomenon was mainly caused by the imbalance between environmental protection and economic growth due to rapid economic expansion [1]. From a spatial perspective, there was an obvious spatial gradient difference in the synergistic effects in industrial green production, showing a spatial distribution characteristic of decreasing from east to west. The first critical cause was that the eastern region had a highly industrialised structure, which promoted the synergistic agglomeration of production in high-value chains compared to the western region. The second critical cause was that the eastern region had a higher endowment of factors such as innovation factors, human capital, green finance, and infrastructure [47,61], resulting in a high potential for synergistic effects in industrial green production. The final reason was that environmental policy implementation in the eastern region was efficient, with a rapid response to green production [60,62].

3.3. Spatial Patterns of Synergistic Effects in Industrial Green Production

3.3.1. Global Autocorrelation Analysis

Table 4. Global Moran’s I index of synergistic effects in China’s industrial green production.

Year	Moran’I	Z (I)	p
2012	0.094	3.622	0.000
2013	0.095	3.651	0.000
2014	0.096	3.666	0.000
2015	0.090	3.530	0.000
2016	0.096	3.653	0.000
2017	0.089	3.452	0.001
2018	0.090	3.514	0.000
2019	0.091	3.557	0.000
2020	0.084	3.389	0.001
2021	0.088	3.509	0.000
2022	0.087	3.501	0.000

shows the global Moran’s I index of the synergistic effects in China’s industrial green production. The results indicate that the values of the global Moran’s I index were greater than 0 in the period of 2012–2022. The p-values passed the hypothesis test of a significance of less than 1% and the Z-values were greater than 1.96, which indicated that the spatial correlation of the synergistic effects in industrial green production among provinces was significant and positive. Moran’s I index showed a small fluctuation trend, indicating that the spatial pattern of the synergistic effects in China’s industrial green production was relatively stable.

3.3.2. Local Autocorrelation Analysis

The global autocorrelation tests had difficulty revealing regional heterogeneity characteristics. Therefore, the spatial aggregation characteristics of the synergistic effects among the industrial green production subsystems were further analysed through changes in the Moran scatterplot, with the results presented in Figure 3. The years 2012, 2017, and 2022 were selected as observation years for the Moran scatterplot. In terms of aggregation type, the synergistic effects in industrial green production were manifested mainly as ‘high-high’ (H-H) and ‘low-low’ (L-L) aggregation patterns, indicating that the synergistic effects in industrial green production between neighbouring regions interacted with each other. In terms of the changes in the types of aggregation, the number of provinces with H-H aggregation increased from 8 to 12, indicating the significant spatial mutual reinforcement of the provinces. Cross-regional synergistic effects in industrial green production need to be promoted in the future, thus allowing H-H agglomerations to drive the development of neighbouring regions.

3.3.3. Directional Distribution and Centre of Gravity Migration Characteristics

The standard deviation ellipse and the centre of gravity migration for the synergistic effects in industrial green production are shown in Figure 4. The coordinates of the centre of gravity for the synergistic effects in the industrial green production of China from 2012 to 2022 ranged between 113°05′ E~113°31′ E and 33°30′ N~33°78′ N. On the influence of the industrial transfer policy [63,64], the centre of gravity migrated significantly to the southwest, with a total migration distance of 74.66 km. The centre of gravity was located mainly in Pingdingshan city and Nanyang city in Henan province. The standard deviation ellipse of the synergistic effects in China’s industrial green production from 2012 to 2022 essentially showed a distribution pattern of ‘northeast–southwest’. The area of the standard deviation ellipse showed a decreasing trend, indicating that the synergistic difference tended to decrease. With the long axis becoming shorter and the short axis becoming longer, the azimuthal angle of the standard deviation ellipse showed an upwards trend, indicating that the synergistic effects in industrial green production were characterised by northeast–southwest centripetal aggregation and southeast–northwest centrifugal aggregation.

3.4. Spatially Heterogeneous Driving Factors for Synergistic Effects in Industrial Green Production

The spatial pattern of synergistic effects in industrial green production revealed the apparent non-equilibrium distribution characteristics across regions. However, this phenomenon of spatial heterogeneity stemmed from the spatial interaction of multiple driving factors [36,37,38]. Therefore, this study further analysed the driving factors behind the spatial heterogeneity of synergistic effects in industrial green production to deeply dissect the driving mechanisms behind the spatial patterns.

3.4.1. Influential Factor Selection and Testing

From the perspective of the industrial ecosystem, the three major factors of basic support conditions, the institutional environment, and the market environment profoundly shaped the spatial heterogeneity pattern of the synergistic effects in industrial green production through nonequilibrium interactions. First, basic support conditions such as green innovation, the level of transport infrastructure, the level of transport infrastructure, the labour level, and the level of economic development influenced the potential for the synergistic effects in industrial green production by providing guarantees like labour, financial, material, technological, and information guarantees [48]. Second, the institutional environment, such as environmental regulations and green credit, improved the synergistic effects in industrial green production by constraining and guiding industrial behaviours [60,65]. Finally, the level of social consumption, the intensity of foreign direct investment, the urbanisation level, and the industrial structure determined the development of the market environment, which influenced the synergistic potential of green production through the efficiency of market resource allocation [66]. Therefore, the driving mechanism of the synergistic effects in industrial green production was systematically explored from the basic support conditions, the institutional environment, and the market environment, with 11 driving factors being screened out.

An OLS model was used to conduct multiple covariance tests for all standardised driving factors, excluding variables with variance inflation (VIF) factors greater than 10 and those that did not pass the significance test. Five driving indicators, namely, the level of digital economy development (DEDL), the intensity of environmental regulation (ERI), the level of green credit (GCL), the urbanisation level (UR), and the optimisation of industrial structure (ISO), were finally filtered out. The relevant descriptions of the driving indicators and OLS regression results are shown in

Table 5. Driving indicator correlation description and OLS regression.

Category	Variable	Description	Data Source	Coefficient & Significance	VIF
dependent variable	−	Synergistic effects of the three phases of industrial green productive capacity	Calculated from Equation (5)	−	−
Constant	−	−	−	−0.607 (−5.60)	−
Basic support conditions	Green Innovation (Ginnov)	Logarithmic number of green patent applications	Chinese Research Data Services (CNRDS)	0.008 (1.33)	31.71
	Level of transport infrastructure (Tinfra)	Logarithm of road mileage	China Statistical Yearbook	0.003 (0.42)	8.72
	Level of digital economy development (DEDL)	Logarithm of the count of terms related to the digital economy	Government working report	0.551 *** (15.20)	3.15
	Labour level (Labour)	Logarithm of total employed persons	China Statistical Yearbook	0.092 *** (7.31)	18.17
	Per capita GDP (PGDP)	Per capita GDP	China Statistical Yearbook	−0.006 (−1.32)	17.02
Institutional environment	Environmental regulation intensity (ERI)	Logarithm of the count of terms related to the environment	Government working report	0.093 ** (2.10)	4.85
	Level of green credit (GCL)	Green credit ratio	The People’s Bank of China	0.058 ** (2.49)	2.52
Market environment	Level of social consumption (Sconsum)	Total consumption of consumer goods/GDP	China Statistical Yearbook	−0.071 (−1.58)	1.57
	Intensity of foreign direct investment (FDI)	Total trade in goods and services/GDP	China Statistical Yearbook	0.006 (0.29)	6.38
	Urbanisation level (UR)	Urban population/total population	China Statistical Yearbook	0.388 *** (7.29)	7.64
	Optimisation of industrial structure (ISO)	Tertiary industry value added/GDP	China Statistical Yearbook	−0.045 *** (−10.52)	1.97

Note: T-values in parentheses; ** p < 0.05; *** p < 0.01.

.

3.4.2. Spatial Heterogeneity Analysis of Impacts of Driving Factors

As shown in

Table 6. Estimates of fitting parameters for OLS and GWR models in 2022.

Model	R2	Adj.R2	AICc
OLS	0.834	0.806	−87.498
GWR	0.932	0.896	−105.100

, compared with the OLS model, the GWR model had the smallest AICc value of −105.100 and the largest adjusted R² of 0.896, so the GWR model had more explanatory power. Therefore, the GWR model was more appropriate to be used in this study to analyse the driving factors of synergistic effects in industrial green production.

The descriptive statistics for the regression coefficients of the GWR model are shown in

Table 7. Descriptive statistics for the regression coefficients of the GWR model.

Variable	Mean	SD	Min	Max	P50	CV
DEDL	1.093	0.901	0.159	3.702	0.956	0.825
ERI	−0.037	1.474	−3.118	3.615	−0.181	−40.295
GCL	0.046	0.897	−3.037	2.244	0.008	19.533
UR	0.799	0.877	−0.154	4.719	0.607	1.098
ISO	0.140	0.155	−0.039	0.746	0.136	1.104

. The strength in the impacts of each driving factor on the synergistic effects in industrial green production was, in descending order, DEDL (+), UR (+), ISO (+), GCL (+), and ERI (−). The absolute values on the coefficients of variation (CV) for their regression coefficients were greater than 0.5, indicating that there was strong spatial heterogeneity in the impacts of the driving factors on the synergistic effects in industrial green production.

The visualised spatial distributions of the GWR results are shown in Figure 5. The results showed that the regression coefficients had good significance, and the GWR model rejected the null hypothesis that the local regression coefficients of the driving factors in specific geographical locations were not significantly different from the global regression coefficients. Therefore, the influence of each driving factor on the synergistic effects of industrial green production exhibited significant spatial differentiation characteristics. The details are analysed below:

(1)

DEDL was positively correlated with the synergistic effects in industrial green production. The extreme values of positive correlation occurred in coastal areas of southeast China, such as Guangdong province, Jiangsu province, Zhejiang province, Fujian province, and Shanghai municipality. However, northeast China, such as the Inner Mongolia autonomous region, Liaoning province, Jilin province, and Heilongjiang province, showed weak positive correlations. This phenomenon was primarily attributed to the more advanced digital economy infrastructure in China’s southeastern coastal regions, where digital technologies were deeply embedded throughout the entire industrial chain [67]. This integration facilitated more effective coordination and control of industrial green production processes. However, in northeast China, the poor digitalization infrastructure increased the cost of digital transformation for industrial firms and inhibited the spread of digital technology adoption [68]. Moreover, the region’s industrial structure, dominated by energy-intensive industries, had low adaptability to digital green technologies, further weakening the role of the digital economy in promoting the green transformation of industrial production [61].

(2)

With respect to ERI, 60% of China’s regional ERI was negatively correlated with synergistic effects in industrial green production. The extreme values of negative correlations occurred in areas with ‘strong provincial capital’ development patterns, such as Sichuan province and Anhui province. The ‘resource siphoning’ phenomenon in these regions concentrated high-quality resources, such as population, capital, and technology, in the provincial capitals, weakening the technological upgrading capacity of industrial enterprises in non-provincial capitals to cope with environmental regulations. Furthermore, provincial capitals may transfer polluting industries to their neighbourhoods in pursuit of economic performance, ultimately leading to a weakening of the overall synergistic effects [69]. Positive correlations were found in economically strong provinces such as Jiangsu province, Zhejiang province, Guangdong province, and Henan province. To enhance economic competitiveness, industrial enterprises in these regions were quicker to respond to synergistic policies [62,70], especially in Zhejiang province and Guangdong province, where the private economy was well developed.

(3)

GCL was negatively correlated with the synergistic effects in industrial green production in northeast China and southwest China, but positively correlated with the synergistic effects in industrial green production in the coastal regions of the East China Sea and South China Sea. This phenomenon stemmed mainly from the fact that northeast China and southwest China relied heavily on traditional heavy industries. Such traditional industries were generally characterised by high asset specificity and high costs of technological transformation [71], resulting in constraints on the synergistic potential of the green credit to support industrial green production. With a high efficiency in the allocation of green credit resources to industrial enterprises, the regions of the East China Sea and South China Sea had a concentration of high technology industries and advanced manufacturing industries, resulting in strong synergistic effects of green production [65].

(4)

A higher UR indicated a stronger ability of the regional population, capital, and technology agglomeration. UR was positively correlated with the synergistic effects of industrial green production for 90% of China’s regions. First, UR effectively reduced the marginal environmental governance cost per unit of industrial output by promoting the centralised construction of environmental protection infrastructure and the large-scale treatment of pollutants in industrial enterprises [66]. Second, the resource agglomeration effect generated by UR made clean technology diffusion more economical, thus reducing the cost of industrial green production [21]. Finally, the increased innovation factor density brought about by urbanisation accelerated green technology spillovers [21], which reduced the cost of industrial green transformation to further drive the synergistic effects of industrial green production [66].

(5)

ISO determined the efficiency of resource allocation in industrial production [61,72]. The extreme values of the positive correlation between this driving factor and the synergistic effects in industrial green production were found in provinces with more marketization, such as Guangdong province, Jiangsu province, and Zhejiang province. The reason was mainly due to the high level of ISO in these regions, which facilitated the development of numerous high-value industrial chains, such as the electronics industry in Guangdong province, the high-value equipment industry in Jiangsu province, and the digital security industry in Zhejiang province. These high-value industrial chains promoted cross-sectoral technology diffusion and drove the optimisation of the industrial ecological circulation system, thereby enhancing the synergistic effectiveness of the industrial green production system [73]. Regions with weak negative correlations between ISO and the synergistic effects in industrial green production were provinces with deeply locked-in energy structures, such as Shanxi province, Liaoning province, Heilongjiang province, and the Inner Mongolia autonomous region. These regions had high degrees of coal dependence and high barriers to energy conversion costs, for which industrial green production was more inclined towards passive compliance than active synergy [74]. In addition, industrial enterprises in these regions had long remained in the high energy consumption segments of the industrial chain, which had a dampening effect on the synergistic effects of industrial green production.

4. Discussion

Previous studies have demonstrated the importance of industrial synergistic effects for sustainable development [16,44] and recognised spatial differences in environmental governance [49,53]. However, existing studies have focused on the synergistic effects between industrial development factors at the macro level. Due to the lack of an analytical framework for synergistic effects among industrial production stages at the micro level, it is difficult to reveal the internal mechanisms of synergistic effects in green industrial production. Therefore, a three-stage synergistic framework for green industrial production was innovatively proposed in this study, providing a unique research perspective in the field of industrial synergy.

Unlike traditional studies that examined industrial green production from one single dimension, such as energy efficiency [23] or pollution control [9,26], this study investigated green production throughout the whole life cycle of SGPMS, PGPOS, and EGPCS from multiple dimensions. Furthermore, the interactions among the three stages were confirmed to generate systematic synergies. The research findings on spatial patterns indicated an increasing trend of synergistic effects in industrial green production, accompanied by significant spatial clustering patterns. This finding aligned with Meng et al.’s [32] arguments regarding regional heterogeneity in environmental performance, illustrating the necessity of regional green production synergy in promoting sustainable development.

The analysis of spatial heterogeneity brought new insights. Liu et al. [42] emphasised that environmental regulation was the dominant determinant of green development. Hao et al. [18] identified technological innovation as the critical factor in the green transition. However, our study showed that DEDL and UR were the major driving factors of synergistic effects in industrial green production, followed by ISO, GCL, and ERI, revealing a more complex pattern of environmental governance. Differences in the ranking of influencing factors were primarily due to the following reasons. First, the digital economy significantly enhanced the allocation efficiency of industrial production factors through the Internet platform, alleviated information asymmetry, and increased resource mobility [8,20]. Second, the agglomeration of innovation factors brought about by urbanisation created a scale effect that surpassed the impact of a single technological innovation [21]. Finally, the relatively low impact of environmental regulation reflected the limitations of policies in promoting synergistic effects among systems, especially in areas with weak institutional enforcement capacity [41,42]. Therefore, the synergistic development of industrial green production needs to build a multi-dimensional promotion mechanism to break through the path dependency of the unidimensional approach.

5. Conclusions and Implications

5.1. Conclusions

To evaluate the green production capacity of subsystems in the three stages of industrial production from a full life cycle perspective, an index system for China’s industrial green production was constructed. Based on the synergetic model and the GWR model, the spatiotemporal patterns of synergic effects in industrial green production were analysed. Finally, the driving factors of synthetic effects in industrial green production and their spatial differentiation characteristics were revealed. The main conclusions were as follows:

(1)

Industrial green production capacity increased rapidly over time from 2012 to 2022, and in terms of regional distribution, it was highest in the east and lowest in the west. Industrial green production capacity exhibited the characteristic of EGPCS > PGPOS > SGPMS. The synergistic effects among the stages of industrial green production clearly increased, shifting from mildly dysfunctional to hardly dysfunctional, with a regional distribution characterised by a decrease from east to west.

(2)

The synergistic effects in industrial green production had a significant positive spatial correlation. The number of provinces with H-H aggregation increased during the study period. The standard deviation ellipse of the synergistic effects in China’s industrial green production essentially showed a distribution pattern of ‘northeast–southwest’. The centre of gravity migrated to the southwest and was distributed mainly in Pingdingshan city and Nanyang city in Henan province.

(3)

Among the five driving factors, DEDL, ERI, GCL, UR, and ISO, DEDL had the greatest influence on the synergistic effects in industrial green production, and ERI had the least influence. The influence of each driving factor on the synergistic effects in industrial green production showed significant spatial differentiation characteristics. The spatial heterogeneity was most prominent for the impacts of ERI on synergistic effects in industrial green production, while the impacts of DEDL had the least differences across regions.

5.2. Implications

(1)

This study confirmed that green production capacity was strongest in EGPCS but weakest in SGPMS. Therefore, the Chinese government should focus on industrial products with outstanding green attributes and high consumption, and formulate green design evaluation standards to guide the Chinese industry in implementing green design and procurement. For industrial enterprises, the concept of green production should be institutionalised. Setting a red line to constrain energy consumption, pollution control, and emissions may affect economic benefits in the short term, but from a development perspective, the full implementation of green manufacturing not only creates new market opportunities for enterprises, but also helps to enhance the brand value of enterprises.

(2)

The number of regions with H-H aggregation for synergistic effects in industrial green production increased and the differences between regions decreased. However, the synergistic effects of industrial green production still remained in a hardly dysfunctional recession type. Therefore, at the industrial enterprise level, sectoral, industrial, and regional constraints should be broken down to synergistically promote innovation in green production technologies through information sharing, system sharing, and experience sharing. At the regional level, industrial chains and industrial parks should be leveraged to strengthen the synergistic effects of green production at the whole life cycle stages. For provinces with better synergistic development of industrial green production, on the one hand, they should continuously explore innovative measures for green production synergy, and on the other hand, they should link up with provinces with lower synergistic development to jointly improve the institutional system of synergistic development for industrial green production. For provinces with low synergistic development of green production in industry, they should actively learn from provinces with high synergistic development and improve synergistic measures to better achieve pollution reduction.

(3)

The driving factors DEDL, ERI, GCL, UR, and ISO revealed the enhancement mechanism of synergistic effects in industrial green production. Based on the spatially heterogeneous influences of the drivers, the Chinese government needs to adopt synergistic measures of green production tailored to local conditions. Eastern regions such as Guangdong, Jiangsu, and Zhejiang provinces should focus on promoting the integration of digital technologies into advanced manufacturing systems and optimising the differentiated pricing system for green credits based on environmental performance. Central regions, like Anhui, Shanxi, and Heilongjiang provinces, need to improve incentive and constraint mechanisms for environmental regulation while establishing restrictive industry catalogues to promote the transition of industrial enterprises. Western regions, especially Liaoning province, Inner Mongolia autonomous region, and Qinghai province, should focus on the matching degree of the urbanisation process with the ecological carrying capacity, strengthen the construction of digital economy infrastructure, to promote the green and intelligent transformation of traditional industries.

5.3. Limitations and Research Prospects

This study has several limitations. First, owing to the availability of data, we constructed an indicator system of industry green production at the overall industrial level. However, there may be differences in the production process in different industrial sectors [13]. Future research will gradually build a system of green production indicators for specific industrial sectors. Second, in this study, we analysed the spatial heterogeneity of the driving factors of the synergistic effects in industrial green production, but the interaction mechanism of the driving factors needs to be further addressed. Finally, this study discussed the spatial heterogeneity of the driving factors for synergistic effects in industrial green production, but the impacts before and after the dual carbon policies may differ [75]. So, how dual carbon policies can improve the synergistic efficiency of the three stages needs to be further explored. In the future, we should deepen the research on the synergistic effects of green production in specific industries and further explore the principle and timeliness of the interaction of driving factors in order to optimise the synergistic path of industrial green production.