From Solo to Cluster Governance: An Empirical Study of Transforming Rural Management in Guiyang, China

Abstract

As China shifts from urbanization to rural revitalization within its rural governance strategy, devising appropriate governance programs becomes crucial for the effective implementation of overarching strategies. This paper explores the policy pathway of the rural revitalization strategy via the lens of village relational governance. This paper builds a relational network of village governance using the Newtonian gravity model and proposes an effective relational governance policy by analyzing the impact of village cluster patterns under different policy rules. Empirical research was conducted in Guiyang County, Hunan Province, China. The findings of this paper are as follows. (1) Rural development in Guiyang County heavily relies on location advantages and natural resources, and there is an urgent need to reinvent the path of rural governance to foster potential rural clusters. (2) A comparative analysis of the relational networks shows that the assignment-based network has more high-clustering groups and fewer low-clustering clusters than the merit-based network, and it has more cluster types, resulting in a more balanced and diverse network structure. In contrast, the merit-based network has fewer cluster types and tends to have a centralized structure. (3) The assignment-type network has greater advantages in terms of agricultural productivity, preserving local culture, and protecting the environment. Simultaneously, its network path has the potential to boost the intrinsic vitality of rural areas and attract more groups to contribute to its development. This path is feasible due to the high level of organization within the Chinese villages. Consequently, this study recommends that the county government should actively decentralize power to villages and grant villages equal development rights to encourage villages to build network clusters with unique competitive advantages.

1. Introduction

Throughout the process of global modernization, the transformation of China’s governance model is an essential chapter that cannot be overlooked, particularly the evolution of rural development programs [1]. From 1978 to 2000, China’s rural policies were primarily focused on increasing agricultural output efficiency and tacking food shortages [2]. From 2001 to 2017, a period marked by rapid urbanization, the government advocated a development strategy for a new form of urbanization that prioritizes people [3]. The rapid growth of urbanization has brought about many positive changes, including reshaping economic and geographic patterns as well as modifying the social governance structure and people’s way of life. However, the process of urbanization has led to many rural issues, such as the “hollowing out” phenomenon caused by the massive outflow of labor [4], and the persistent marginalization of rural areas in the national development strategy, which has resulted in insufficient resource allocation and policy support [5]. In recent years, as China’s urbanization rate approaches 60% and begins to slow down [6], rural concerns associated with urbanization have emerged. To further promote comprehensive modernization, the Chinese government has incorporated “Rural Revitalization” as a major strategy in national strategic planning [7]. This strategy is not only a response to the progression of urbanization to a certain point but also represents a significant effort to map out a new path for modernization with Chinese characteristics.

The goal of the rural revitalization strategy is to activate the endogenous dynamics and promote holistic advancement in rural areas, thereby building a more equitable social structure [8]. The strategy is not implemented in isolation but rather integrated into China’s distinct governance system. In this governance system, county governments act as limited liability managers, allocating resources by selecting typical villages, implementing policies, such as new rural construction and traditional village protection, and encouraging other villages to emulate the typical villages [9]. However, while this type of management has, to some extent, facilitated village development and change, it has also given rise to emerging challenges over time. County governments tend to promote rural governance as a specific task by selecting a few villages for resource investment through the issuing targets and failing to form a systematic institutional arrangement and design. This piecemeal approach to governance, focusing on individual villages, leads to fragmentation within the county’s rural areas, resulting in low participation and motivation in other villages. This not only hinders the in-depth implementation of the rural revitalization strategy but also further exacerbates the problem of fragmentation and disorder in rural governance.

Existing literature has addressed various dimensions of rural governance. Among them, some scholars have proposed a macroscopic governance framework that delves into the structure, functions, and operational mechanisms of rural areas [10,11], and such research can help to shape governance policies. The research methodologies employed by researchers for rural governance study include spatial analysis [12], social network analysis [13], cluster analysis [14], and comprehensive assessment [15], etc., which provide valuable tools for analyzing the spatial scope and object of rural governance. Some scholars have addressed issues in rural governance, such as spatial fragmentation, inefficiency, and substantial inter-regional differences [16], and they have explored whether collaborative governance can offer specific benefits for rural development and provide a pathway for rural governance [17]. Additionally, there are studies on the problems that arise during the formulation and implementation of governance policies [18,19], and resolving these issues can provide policy-level direction for rural governance. However, although significant advancements, there are still certain shortcomings in prior research. Firstly, these studies predominantly focus on macro-level or theoretical aspects, with limited discussion of specific governance programs and their effects and a scarcity of research cases on typical counties. Secondly, in terms of methodology and policy research, policy suggestions primarily focus on village unit governance, while there is a lack of analysis on governance policies for rural clusters. Although some studies have suggested that relational governance, as a new governance model for villages, might stimulate the overall development vitality of villages and increase village involvement by strengthening village collaboration [20]. Regarding existing research on collaborative village governance, scholars have primarily conducted preliminary research on village cluster governance, exploring horizontal scale clusters or vertical industrial agglomerations along the supply chain, which eventually self-organize into localized clusters with a small number of villages [21]. There is a dearth of comprehensive studies and a clear direction on the types of rural clusters that should be developed, as well as the effective policy mechanisms needed to promote the growth of village clusters.

Villages cannot flourish in isolation; they must be interconnected and developed within a broader organizational system. Therefore, the goal of this study is to effectively organize villages through relational governance and to fully capitalize on the beneficial co-opetition relationship among villages in governance. The contributions of this study are as follows: (1) On the theoretical level, it explores the advantages of inter-village relational governance over traditional village unit governance, emphasizing the importance of fully utilizing the beneficial co-opetition relationships among villages in county-level governance. (2) At the methodological level, this study utilizes social network analysis to simulate the development results of rural clusters under different policy scenarios, assesses their effects, and examines the types of policies needed to promote cluster growth, thereby filling a research gap in cluster policy support methodologies. (3) This study is closely associated with a rural survey of typical counties, collects firsthand data, empirically tests research hypotheses and theoretical models, and offers a realistic basis for policy design and implementation.

This paper is structured as follows. First, the analytical framework and research methodology are systematically described; The inter-village co-opetition relations are clearly defined, and the methodologies for constructing the relationship network under meritocratic and assignment policies, as well as the analytical approach, are identified. Secondly, the study area, Guiyang County in Hunan Province, China, and the data sources are described in detail. The aforementioned analytical methodologies are employed to thoroughly examine the structural characteristics and comparative advantages of the two models of rural relational networks in Guiyang County, revealing the distinctions and impacts of the two models through comparative analysis. Finally, the key findings of this study are presented, and policy recommendations for rural governance are offered in light of these findings.

2. Theoretical Framework

In a stable rural society, a complex network of relationships develops among villages due to resource competition and mutual production assistance. This network gradually takes on a stable internal organizational structure as it evolves through resource competition and integration, manifesting as village clusters [22]. It has numerous benefits, including enhanced village cohesion and attracting investment, thereby laying the groundwork for rural spatial development projects [23]. The cluster formation in villages is driven by underlying logic, and its organizational process can be bottom–up or a combination of top–down and bottom–up, requiring the assistance of industry, policy, villagers’ self-governance, and financing [24]. Village revitalization aims to promote the prosperity of villages and improve the living conditions of their residents [25]. Village cluster governance is based on village organizational structures and aims to achieve long-term economic, social, and environmental development in rural areas by integrating regional resources, optimizing spatial layout, enhancing industrial synergies, and improving governance capacities [26]. From the perspective of policy governance for cluster formation, this study explores which type of policy tool is most effective in shaping the optimal village cluster mode, thereby offering a policy pathway for the implementation of rural revitalization strategy.

Beginning with the governance of inter-village relationships, firstly, the co-opetition relationships between villages are measured and form a network of all relationships in the villages. When investigating the relationship between villages, it is found that these relationships are influenced by the development capability of each village and the spatial distance between them [27], akin to the Newtonian gravity calculation theory. Consequently, this study employs the Newtonian gravity model to assess inter-village relationships. Specifically, the assessment of a village’s development capability considers the four primary Sustainable Development Goals (SDGs): geography, economy, population, and administration. Secondly, the characteristics of relationship networks are obscured by the gravitational model, which contains a huge number of gravitational forces. To distinguish between primary and secondary relationships in the network, the key gravitational network was constructed by selecting edges with higher gravitational forces in the network (based on merit) and distributing equal gravitational forces to each village (based on assignment). In other words, this is achieved by simulating two different scenarios: one where power is allocated based on the county government’s merit-based selection, and another where power is distributed according to the villages’ equitable development. Multiple sets of experiments are carried out to guarantee the scientific validity of the findings. Finally, as the village cluster groups vary greatly under different policy rules, this study compares the structural characteristics of various network modes using social network analysis. The spatial and typological characteristics of village clusters are analyzed using the network composite index and Moran’s I, evaluated across sustainability perspectives, and this information is then used to refine the policy rule for the clusters (see Figure 1).

This study aims to encourage the county to establish reasonable policy regulations to support the establishment of village clusters through relational management strategies, thereby mobilizing potentially developing small communities into a village commonwealth society.

3. Methodology

3.1. Rural Sustainable Development Capability Assessment

The assessment of village’ sustainable development capability determinates its co-opetition relationship. This study combines the economic, environmental, and social goals of sustainable development with empirical research data to create rural sustainable development capability assessment system that considers the geographic environment, economic vitality, population dynamics, and administrative management [28,29]. These four dimensions influence the capacity and sustainability of village development. The geographical environment encompasses natural resources, such as land, water, and plants, which are the foundation of village development and guarantee ecological security [30]. The economic dimension focuses on the industrial component, which can increase villager’s income, create employment opportunities, and revitalize rural areas [31]. The population dimension constitutes rural society and determines the sustainability of village habitation [32]. Administrative factors foster greater inter-village collaboration by establishing a more peaceful, equitable, and resilient governance structure [17].

Adhering to the principles of scientific rigor, feasibility, systematicity, and representativeness, and based on the four core elements of geography, economy, population, and politics, as well as the existing literature on the sustainable rural development assessment [28,33], this study identifies indicators that can adequately reflect village’s sustainable development capacity. Experts in specialized domains are also consulted to screen these indicators. Specifically, representative indicators of county administration, such as infrastructure development and policy support, are sifted to ensure their relevance. Furthermore, essential indicators for assessing rural sustainable development capacity should be selected based on data accessibility, generalizability, and comparability. The resulting evaluation system comprises four first-class indicators, nine second-class indicators, and twenty-three third-class indicators, as detailed (see

Table 1. Summary of rural sustainable development capacity indicators.

First-Class Index	Second-Class Index	Third-Class Index	Explanation
Geographical environment	Terrain and landform	R1 Terrain relief	Elevation range within a village
		R2 Surface fragmentation	The condition of surface incision density and incision depth within the village
	Location	R3 Average distance to the county	Average distance from village center to county government
		R4 Average distance to the township	Average distance from village center to township government
		R5 Average distance to traffic trunk line	Average distance from village center to main roads above provincial level
	Ecological resources	R6 Proportion of ecological land area	Ecological land area/total area of village
		R7 Average distance to the river	Average distance from village center to the nearest river
Economic vitality	Agricultural industry	R8 Proportion of cultivated land area *	Cultivated land area/total area of village
	Industrial and commercial industry	R9 Proportion of commercial land area *	Commercial land area/total area of village
		R10 Number of village enterprises *	POI data statistics
		R11 Proportion of industrial and mining land area *	Industrial and mining land area/total area of village
Population dynamics	Total population	R12 Population size *	Village population size
		R13 Population density *	Population size/total area of village
	Population movement	R14 Amount of population movement	Number of villagers working externally
		R15 Amount of population loss	Number of villagers who have relocated
Administrative management	Township governance	R16 Infrastructure availability *	The availability of infrastructure such as roads, water, electricity
		R17 Public service availability *	The availability of public services including education, health, culture
		R18 Agricultural policy support *	Township support for agricultural development through technology, subsidies
		R19 Industrial policy support *	Township support for agricultural industries
		R20 Environmental governance *	Township financial investment in village environmental governance
		R21 Public security	Incidence of security issues in a village
	Village autonomy	R22 Participation of village assemblies *	Participation rate in village assemblies
		R23 Number of rural elites *	Number of village committee members, entrepreneurs, technical experts, teachers

Note: indicators with * have a positive contribution.

).

The composite index method is used to evaluate the sustainable development capacity of villages. Firstly, the indicators are normalized and converted to [0, 1] interval according to Formula (1) for positively correlated indicators of rural sustainable development capability. The negatively correlated indicators are translated into [0, 1] interval using Formula (2).

$$u={\displaystyle \frac{y-y_{min}}{y_{max}-y_{min}}}$$

(1)

$$u={\displaystyle \frac{y_{max}-y}{y_{max}-y_{min}}}$$

(2)

where u denotes the standardized data of a certain indicator, y is the original value, and $y_{max}$ and $y_{min}$ are the maximum and minimum values, respectively.

Secondly, indicators at various levels are weighted and summed, with the weights determined through expert opinions.

$$Q_i=\sum w_i{u}_i(i=1,2,3,\dots ,23)$$

(3)

where $Q_i$ denotes the sustainable development capability of village i; $u_i$ denotes indicator i; $w_i$ denotes the weight of indicator i.

3.2. Gravity Model

Contrasting with the traditional governance model that categorizes villages, this study focuses on managing inter-village relationships to optimize rural regeneration pathway. The co-opetition relationship between villages needs to be measured. When investigating the relationship between villages, it is found that it is influenced by the development ability of villages, robustly developing villages tend to forge stronger interactive bonds with their neighbors. It is also influenced by the first law of geography, which states that the closer the distance is, the more similar the form is, akin to the Newtonian gravity calculation theory. The gravity model has been integrated into geographic studies and widely employed in urban spatial analysis and rural settlement research [34,35], demonstrating its scientific rigor and validity. When assessing the strength of inter-village relationships, the gravity model offers a scientifically sound approach. In the model, “quality” represents sustainable development capability of a village, while “distance” reflects the spatial distance between villages and the closeness of their connections.

$$F_{ij}={\displaystyle \frac{Q_i\times Q_j}{D_{ij}}}$$

(4)

where $F_{ij}$ signifies the gravitational force between village i and j; $D_{ij}$ represents the distance between village i and j; $Q_i$ and $Q_j$ represent the sustainable development capability of village i and j, respectively.

Gravitational relationships among villages have created a complex virtual network in the county.

$$N=m\times (m-1)/2$$

(5)

N is the amount of network gravity and m is the number of villages.

The complicated virtual network contains primary and secondary gravitational relationships, and in order to reflect the network’s characteristics, it is vital to identify the primary gravitational relationship. Complex network theory suggests two methods for selecting primary gravity: selecting the gravity with the higher value and allocating an equal amount of gravity to each village according to the principle of fairness [36]. These methods represent different policy options regarding whether to decentralize power to villages. Since the number of gravitational edges has an impact on the network structure and the characteristics of village clusters, multiple experimental sets are conducted in this study to ensure scientific results. Merit-oriented networks are constructed by 5%, 10%, and 20% of the total gravitational edges [37], which illustrate the effects of a few, some, or many gravitational ties on the network structure. The assignment-type networks are constructed by the maximum, the top 5, and the top 10 gravity edges in a single village to reflect the effect of a few, some, or many options on the network structure, respectively. Different simulation settings generate different network structures and have different impacts on village development. Therefore, this study optimizes the governance pathway by comparing the characteristics and impacts of different network structures and feeding back to the policy scenarios.

3.3. Social Network Analysis

Social network analysis (SNA) is a network structure analysis method that can clearly display the position of nodes in the network [38], assess the complicated network structure quantitatively, and extract key structural information. Because the network generated by inter-village gravity is quite complicated, this research employs SNA to analyze the characteristics of the gravity network formed under different policy rules. To ensure the study’s systematicity and rigor, the indicators of network analysis should be screened, and this study selects three types of indicators: basic indicators, connection indicators, and clustering indicators, for a comprehensive analysis of the network structure’s underlying characteristics (see

Table 2. Brief description of rural network indicators of SNA.

Indicators	Explanation
Nodes	The number of points in the network
Ties	The number of edges in the network
Density	The actual number of ties is divided by the maximum possible number of ties
Average Degree	The average edges of per node in the network
Connectedness	The connectivity in a network, indicating if the network is a connected graph
Fragmentation	Measuring the number and size of connected components
Compactness	Measuring the closeness of connections between nodes in the network
Average Distance	The average shortest path length between all pairs of nodes in the network
Small Worldness	Indicating whether the network has aggregations with high clustering coefficient and short average path length
Closure	Measuring the number of closed triples in the network
Prop Within Three	The proportion of nodes in the network that can reach each other within three steps
Degree Centralization	Measuring the centralization of the degree distribution in the network

).

To further quantify the spatial characteristics of village clusters in the network, Ties, Average Distance, Density, and Degree Centralization, which are the indicators of the SNA, are chosen to construct a comprehensive index of individual network, which reflects the network characteristic of individual villages. Furthermore, the characteristics of village clusters can be reflected by the index.

$$A_{net\left(i\right)}={\displaystyle \frac{\left(Tie{s}_{\left(i\right)}+Average Distanc{e}_{\left(i\right)}+Densit{y}_{\left(i\right)}+Degree Centralizatio{n}_{\left(i\right)}\right)}{4}}$$

(6)

$A_{net\left(i\right)}$ is the individual network composite index for $Villag{e}_{\left(i\right)}$;$Tie{s}_{\left(i\right)}$, $Average Distanc{e}_{\left(i\right)}$, $Densit{y}_{\left(i\right)}$ and $Degree Centralizatio{n}_{\left(i\right)}$ which denote the values of $Ties$, $Average Distance$, $Density$ and $Degree Centralization$ for $Villag{e}_{\left(i\right)}$, respectively.

3.4. Exploratory Spatial Data Analysis

Exploratory spatial data analysis (ESDA) is a spatial analysis method used to reveal spatial relationships and patterns within data [39]. This paper chooses the Global Moran’s I to reflect the spatial correlation of the individual network composite index. The calculation formula is

$$I={\displaystyle \frac{n{{\displaystyle \sum }}_{i=1}^n{{\displaystyle \sum }}_{j=1}^n{W}_{ij}\left(x_i-\overline{x}\right)\left(x_j-\overline{x}\right)}{{{\displaystyle \sum }}_{i=1}^n{{\displaystyle \sum }}_{j=1}^n{W}_{ij}{{\displaystyle \sum }}_{i=1}^n{\left(x_j-\overline{x}\right)}^2}}$$

(7)

Among them, n represents the total number of villages; x_i and x_j are the values of individual network composite index of village i and j; $\overline{x}$ is the average value of all regions; and W_ij represents the spatial weight matrix. The range of values for the global Moran’s I index is [−1, 1].

The Z-test is commonly used to standardize Moran’s I values for auto-correlation. When the Z-value is positive and significant, it indicates the presence of positive spatial auto-correlation. When the Z value is negative and significant, it indicates the existence of negative spatial auto-correlation. When the Z value is 0, it indicates that the observed values have an independent random distribution.

$$Z\left(I\right)={\displaystyle \frac{I-Q\left(I\right)}{\sqrt{VAR\left(I\right)}}}$$

(8)

Among them, Z(I) is the significant level of Moran’s I. Q(I) is the exponential mean, and VAR(I) is the variance of Moran’s I.

Furthermore, this paper further analyzes the spatial types of individual network composite index under different network modes using local Moran’ s I of ESDA.

4. Study Area and Datasets

4.1. Study Area

Guiyang County is situated between 112°13′26″ and 112°55′46″ east longitude and 25°27′15″ and 26°13′30″ north latitude. Guiyang County is a significant county in Hunan Province under the administration of Chenzhou City (see Figure 2). Covering approximately 2973 square kilometers in total, the county shares borders with Beihu District to the east, Xintian County to the west, Linwu County to the south, and Yongxing County and Leiyang City to the north. It is approximately 33.5 kilometers from the center of Chenzhou, which is a favorable geographical location. Guiyang has a subtropical humid monsoon climate. The altitude ranges from 13 meters to 1422 meters, with its terrain being higher in the north and south and lower in the center. The total population in 2020 is 911,900, including 724,200 permanent residents. The urbanization rate in Guiyang County is 54.52%. It has 22 townships, for a total of 395 villages. The county is representative in numerous ways. The primary economic activity in Guiyang County is agriculture, which yields a variety of products, including rice, roasted tobacco, oil tea, and fruits. The county is also renowned for its nonferrous metal industry and has actively promoted tourism in recent years. Guiyang County has exhibited comparatively robust development in southern China as a result of various agricultural and industrial policies that have led to a certain scale of industrial agglomeration.

However, Guiyang County still faces a series of challenges that are common in rural areas of China. (1) Although the Guiyang County government has chosen numerous characteristic villages for development, their development is ineffective, and the phenomenon of rural hollowing-out is quite severe in Guiyang. (2) Individual farmers continue to dominate agricultural product production, with low added value in the processing industry, indicating that rural industry development is still in its early stages, as well as a lack of policy guidance. (3) Rural industrial development resources are limited, leading to resource competition among villages and a deficiency in coordinated development planning. Therefore, it has become an urgent and significant task to explore the policy pathway of “village clustering” based on rural resource endowment.

4.2. Data Sources

The Natural Resources and Planning Bureau of Guiyang County provided the land use status data, the Guiyang County boundary data, the village boundary data, and the road network data used in this study. Economic and social data, including the population and GDP, were gathered from the Guiyang County Statistical Yearbook (2020) and the Guiyang County Township Statistical Yearbook (2020). Data on enterprises, elementary schools, medical institutes were obtained from the POI data of Gaode Map. The 30m digital elevation model data ASTER GDEM V2 were acquired from the Geospatial Data Cloud (http://www.gscloud.cn, accessed on 1 July 2020). Using land use data as the basic map, various datasets including geographic, POI (Point of Interest), demographic, and economic information were integrated into the projection coordinate system of WGS_1984_UTM_Zone_49N through a GIS platform to establish the Guiyang County Rural Basic Database.

In addition, our research team conducted comprehensive field surveys of Guiyang County’s townships in August 2021 and August 2022. Symposiums were organized to facilitate in-depth discussions with key individuals, such as village heads, accountants, and party branch secretaries, to collect basic data from each village. The gathered data encompasses demographic information, industrial structure data, infrastructure and public service facility status data, agricultural policies, industrial development policies, financial investment in environmental management, village security situations, participation in village assemblies, and the number of elite groups within the villages. From September to October 2022, the village information and data were systematically organized, aiming to comprehensively, objectively, and accurately reflect the characteristics of villages in Guiyang County at this stage.

5. Results and Discussion

5.1. Assessment of Rural Sustainable Development Capacity in Guiyang

The current status of rural development in Guiyang County was evaluated utilizing the results derived from Formula (3). As illustrated in Figure 3, the current development state in the countryside exhibits a spatial distribution pattern, with higher levels observed in the southeast and central regions, gradually diminishing towards the periphery and reaching its nadir in the northern and southern fringes. Regions with high values are primarily located around the county town, encompassing Longtan, Lufeng, Taihe, Zhenghe, and others. These places have a better geographic environment, location conditions, and infrastructure, thus contributing to a higher comprehensive evaluation index. The relatively high-value areas are divided into two segments: one surrounding the perimeter of the high-value areas, benefiting from industrial strengths, and the other situated near the Chongling River, abundant in nonferrous metals and mineral resources, leading to more intensive mining activities. Medium-value zones surround relatively high-value zones, which, while are less developed in infrastructure and industry compared to high-value and relatively high-value zones, posse relatively larger populations and better villages’ autonomy. The low-value areas are primarily distributed in the north and south peripheral areas, characterized by relatively disadvantageous geographical environments and location conditions, predominantly agricultural industries, and inadequate infrastructure. Currently, rural development in Guiyang County is heavily reliant on location advantages and natural resources. However, as the urbanization process comes to an end, it is no longer feasible to rely on urbanization to drive the development of rural areas around the county, and relying on resources to promote rural development will eventually lead to resource depletion. In view of this, it is necessary to explore a new model of rural governance. To ensure the long-term viability of rural development, efforts should be made to tap potential clusters in the rural areas through relational governance.

5.2. Network Characteristics of Village Relationships

5.2.1. Construction of Rural Key Gravitational Networks

The previous section evaluated the current development situation and the governance challenge in county village, and this section analyzed the structural characteristics of the network formed under different policy scenarios from the perspective of village relationship governance. Firstly, the network structure resulting from all gravity forces in Guiyang County village was calculated using Equation (4), which is a complex network containing 73,920 gravitational forces. To enhance the visibility of gravitational characteristics, the calculated results were multiplied by a factor of 10000 in this paper. Secondly, the key gravitational forces in the gravitational network were extracted. Based on the method of constructing merit-based network, three groups of experiments with weak linkage (5%), medium linkage (10%), and strong linkage (20%) were selected, resulting in networks with 3695, 7392, and 14,784 gravitational edges, respectively (see Figure 4). Similarly, using the method of constructing an assignment-based network, three groups with one option, five options, and ten options were chosen, producing networks with 385, 1925, and 3850 gravitational edges, respectively (see Figure 4). The analysis of these two model networks shows that as the number of gravitational edges increases, the network transforms from a sparse, large lattice structure to a dense, complex lattice structure. Given the complexity of the network structure, a thorough analysis of the network characteristics is necessary.

5.2.2. General Characteristics of Rural Key Networks

Some structural parameters of networks were computed using SNA.

Table 3. The description of the rural key gravitational network in Guiyang.

Network Description	Merit-Based Networks			Assignment-Based Networks
	Few Ties	Medium Ties	Numerous Ties	Few Options	Medium Options	Numerous Options
Nodes	385	385	385	385	385	385
Ties	3696	7392	14,784	385	1925	3850
Density	0.05	0.10	0.20	0.003	0.013	0.026
Average Degree	9.60	19.20	38.40	1	5	10
Connectedness	1	1	1	0	1	1
Fragmentation	0	0	0	1	0	0
Compactness	0.263	0.373	0.509	0.003	0.131	0.191
Average Distance	5.35	3.55	2.46	1.42	11.37	7.56
Small Worldness	5.36	3.56	2.33	-	10.76	7.76
Closure	0.62	0.64	0.66	0	0.41	0.54
Prop Within Three	0.270	0.519	0.835	0.004	0.069	0.144

shows that in both networks types, as the number of gravitational edges increases, the Ties and Density increase significantly, the Average Degree also increases dramatically, resulting in a denser network with nodes more closely interconnected. The Compactness increases gradually, while the Average Distance decreases gradually, enhancing network efficiency. In merit-based networks, both Connectedness and Fragmentation are 1 and 0 respectively, indicating a complete connectivity structure at the global level for all networks. The assignment-based network, with its low selectivity, does not yet possess a fully connected structure and exhibits local discontinuities; however, as selectivity increases, the network transitions to a completely connected structure, markedly improving connectivity. Overall, as the number of network edges increases, both the merit-based networks and assigned networks are more compact and efficient.

The values of the Small Worldness, Closure, and Prop Within Three metrics show that the assignment-based networks have a higher number of highly clustered groups and a lower number of lowly clustered cliques compared to the merit-based network. As the number of gravitational edges increases, the number of highly clustered groups with shorter average path lengths gradually declines, whereas the number of less clustered cliques increases in both types of networks. Given that both types of networks exhibit significant clustering characteristics, it is necessary to further analyze the spatial characteristics of village clusters in order to gain a better understanding of these characteristics and their impact on village governance.

5.2.3. Spatial Distribution Characteristics of Village Clusters

The network composite index for each village was calculated using Equation (6), and the scores of composite index were divided into five types through the natural discontinuity grading method in GIS platform, which includes low-value areas, relatively low-value areas, medium-value areas, relatively high-value areas, and high-value areas (see Figure 5). In the merit-based networks, the strong, medium, and weak gravity networks all form bigger villages’ aggregates in the northwest and southeast regions, with clear circle characteristics. As the number of network edges increases, regional circles of the same rank from various aggregates begin to merge. When the weak linkage network shifts to a medium linkage network, the distinction between the two larger aggregates in the northwest region becomes less clear, and the relatively high-value areas seem to expand over a wider area. At the same time, the links between network aggregates in the northwest and southeast areas begin to strengthen, and the high-value regions have greater growth. When the medium linkage network shifts to the strong linkage network, the areas with high value and the relatively high-value areas of the two aggregates in the northwest region gradually increase, and the boundary between the aggregates in the southeast and northwest regions begins to blur, with the relatively low-value areas transforming into a medium-value areas. The two aggregates are further tightened, and the scale of the two larger village clusters in the network continued to increase. Overall, the villages form two core clusters, and as the number of network edges increases, the network aggregates’ cluster power grows, indicating a centralized integration trend.

Meanwhile, in the assignment-type network, as the options of villages have increased, the village gravity network in Guiyang County has steadily evolved into a decentralized system with multiple centers. In the less-selective network, the feature of village clusters is less visible. However, as the less-selective network transitions to the medium-selective network, the outlying portions of Guiyang County begin to form easily identifiable village aggregates. These aggregates have higher comprehensive evaluation values, more gravity edges, shorter gravity lengths, higher densities, and greater node centrality. When the medium-choice network is converted to a multi-choice network, the number of high aggregators reduces, but they retain their dominant position in the network. At the same time, the positions of some network aggregates have undergone turnover, transforming from high-value areas to relatively high-value areas or medium-value areas. However, these areas continue to hold a dominant position in their surrounding circles, occupying the central position of the surrounding circles, and the polycentricity spatial distribution characteristic of village clusters remains significant.

5.2.4. Types of Village Clusters

To further analyze the spatial types of network clusters, this study used Moran’s I to assess the spatial correlation of individual network composite index. The shapefiles were imported into GeoDa 1.18 software, and the global Moran’s I values are significantly positive at the 1% confidence level, indicating a significant positive spatial correlation of all individual networks. To analyze the spatial properties of various individual network, GIS platform was used to map the types of individual network composite index using Local Moran’s I (see Figure 6). The findings demonstrate that individual merit-based networks in Guiyang have significant positive spatial correlation characteristics. The village clusters are divided into high–high and low–low spatial types. The high–high-value areas primarily concentrate in the southeastern areas and the low–low-value places disperse throughout Guiyang County’s northern areas. The types of assignment network are more diverse, being classified into four types: high–high, high–low, low–high, and low–low. Furthermore, the assigned networks have a substantially larger high–high-value area than the merit-based network do. Different village cluster governance strategies are required for different types. The high–high, high–low, and low–high value areas of assignment networks can implement their own village enhancement plans to drive the cluster growth, which will then support the rural overall sustainability. For example, for the high–high areas, the region’s core competitiveness can be enhanced so that it can radiate to the adjacent villages. For both high–low and low-high areas, the inadequacies of relatively weak villages can be compensated for, enabling them to be fully integrated into the realm of high-value areas. However, in the case of merit-based networks, the effective strategies for village governance are more limited, and as they are implemented, they can worsen the demise of more rural places, such as low–low cluster areas.

5.2.5. Effects of Different Rural Cluster Modes

The merit-based network with few ties and the assignment network with numerous options exhibit network structures comprised of 3696 and 3850 edges (see Figure 7), respectively. Due to the close number of edges, their characteristics can be further compared. The merit-based network with few ties forms village clusters that exhibit strong centralization, resulting in fewer clusters overall, with only two major ones. The Territorial Spatial Master Plan (2021–2035) for Guiyang County suggests establishing the county’s northern and southern hubs by leveraging the county town and the Chongling River, respectively. The villages surrounding the county town have developed a strong core region as a result of the concentration of enterprises, such as smart furniture manufacturing, high-tech industries, and technology agriculture. The villages along the Chongling River have also created significant industrial clusters as a result of their advantages in the development of certain industries, such as colored technology and specialty agriculture. The spatial distribution of the village clusters formed by the merit-based network aligns with the government’s development plan, but the actual cluster area is larger than projected. These two clusters can promote the development of villages by utilizing their advantages of industrial resources. However, if resources are concentrated to develop these two areas, the population of the countryside will eventually cluster to the county’s peripheral areas, compounding the problem of rural population decline. Ecological function in city-side areas, as well as agricultural production activities far away from urban areas, may be disappearing. Furthermore, resource-dependent rural development will eventually suffer resource depletion [40], raising the issues of rural long-term sustainability.

In contrast, the assignment-based network with numerous options creates village clusters with a more decentralized and polycentric distribution (see Figure 7). The village clusters in the southeast and west exhibit some overlap with the merit-based network’s village clusters, yet numerous novel village clusters have emerged in the northern and southern areas, most of them are village clusters with agricultural enterprises. In terms of agricultural production, farming activities in hilly areas of southern China, such as Guiyang County, are limited by the radius of cultivation [41], and farmers engaged in agricultural production require some housing and producing facilities nearby, agro-processing industries also need agricultural product cultivation nearby. As a result, cultivating some smaller-scale decentralized village clusters in these locations can better maintain the local agricultural production function. The cultivation of roasted tobacco and oil tea in Guiyang County also has a rich history and culture that can be preserved through village cluster development. The drought and water crop rotation model of rice and roasted tobacco also contributes significantly to improving the local soil microclimate. Furthermore, organizations, such as fruit and tea farmers, have the ability to influence village development in the future [13,20]. By cultivating new tiny clusters, these groups can be encouraged to actively engage in rural construction. The development of the entire rural area of Guiyang can be revitalized via the long-term efforts of numerous generations, which will have a far-reaching impact on rural revitalization. Upon further observation of the village cluster types in the two models, it is found that merit-based network exhibit clusters with high–high and low–low values (see Figure 7). Effective measures in high-value areas may exacerbate the decline of other areas. On the other hand, the merit-based network exhibit four types of village clusters: high–high, high–low, low–high, and low–low. Adopting differential upgrading strategies for different cluster types can drive the sustainable development of a larger number of villages.

5.3. Discussion

Against the backdrop of global rural decline, countries are exploring different models of rural governance, which primarily consist of two models. These comprise the protection of individual villages through the excavation of characteristics, such as village agriculture, architecture, and culture, and the restoration of natural ecological landscapes [42]. This model is geared toward individual village governance and promotes village development by leveraging regional characteristics. However, this model lacks the ability to mobilize development across larger rural areas, and its evaluation criteria can be subject to subjective judgments. The other option is to support the economic growth of villages through the promotion of village industrial clusters [43]. There are two ways of promoting village economic development, horizontal scale clusters and vertical industry mutual aid clusters, both of which focus on the industry’s supply chain and ultimately contribute to a localized area, with no efficient utilization of policy tools to promote the growth of such connections. The establishment of village clusters requires underlying logics, such as industry, policy, villager autonomy, and funding [24], but current research lacks in-depth understanding in this field. The relational governance proposed in this study explores the integration of policy tools into the support system for cluster formation. The results verify that the county’s decentralization of authority to villages promotes the formation of diversified network clusters in villages. Compared with domestic and international village cluster studies based on industries, this study can use policy resources to foster village clusters, which is better suited to the implementation of national rural revival plans. Meanwhile, decentralization has led to diverse village clusters, each cluster with its own resource potential that can be enhanced by policy measures. Compared to previous studies, this study’s findings offer policy tools that will provide a landing spot for additional support.

Although the village cluster structure established by assignment-based networks is more sustainable than that formed by merit-based networks, the feasibility of its path in relational county governance requires further discussion. Chinese villages are highly organized, with each possessing an organizational institution comprising villagers' committees and villagers' groups that oversee the management of village affairs [44]. Villages’ high degree of organization allows them to self-manage and make independent decisions about development. This enables villages to respond rapidly and collectively in the face of external problems and changes, such as COVID-19 prevention management, where the Chinese government implements stringent village-level prevention and control. Although this does impose certain restrictions on the individual freedom of villagers, it notably safeguards the health and safety of the elderly, children, pregnant women and other vulnerable groups with weakened immune systems within the village [45]. In relational governance, the assignment network’s construction should decentralize power to the village collective, providing it with sufficient autonomy to form ties with other villages and support its own development. Because of the village’s strong organization and autonomy, as well as a clear organizational structure for responsibilities, the village can effectively implement county policies and make timely decisions in determining the best path for its own development, thus ensuring policy implementation feasibility for the assigned network. Consequently, the policy implementation path proposed in this study is suitable for rural governance in China.

6. Policy Implications and Conclusions

6.1. Policy Implications

At present, the industrial clusters have initially been shown in rural areas of China, however, numerous potential clusters remain untapped due to insufficient state support and cultivation [46]. To strengthen the foundational logic behind the formation of village cluster, the relational governance method should be clearly articulated within rural governance policies. This paper analyzes the trend of village clusters under different policy rules. Through effect analysis, it concludes that the network mode of empowering multiple options from the county to the villages has obvious comparative advantages in aligning with rural revitalization strategy and preserving village function. The main reason is that such policy support can promote the development of potential clusters in villages, thus allowing villages to choose the most appropriate industrial cluster path for their growth based on resource endowment, comparative advantage, and market demands. Based on this, policy recommendations for rural governance are proposed (see Figure 8).

In the specific path of implementation of rural revitalization, the first task of county governments is overall coordination, which should actively explore the path of relational governance by decentralizing power to villages and giving villages equal rights to autonomous choice and development. During the process of decentralization, guidance for rural development should be strengthened. A rural industrial cluster development fund should be established to provide villages with financial support and risk protection, thereby supporting rural development and attaching importance to healthy competition among villages. On the one hand, this will encourage cooperation and exchange between villages. On the other hand, it will guide the formation of a moderate competition mechanism between villages. Villages should define their development goals and directions in accordance with their own resource endowments, industrial characteristics, and market demands. By exploring their respective advantageous resources and developing special industries, the complementary advantages and synergistic development of different village clusters can be realized. Additionally, villages ought to articulate their individual responsibilities to maintain coherence in the county’s overarching developmental trajectory and guarantee the efficient execution and ongoing monitoring of policies. This will allow for the improvement of supporting measures and mechanisms and timely amendment of relevant laws to provide solid legal safeguards for decentralization. The county government can assess and inspect the development status of rural industrial clusters on a regular basis by way of establishing a monitoring mechanism to identify grassroots problems in a timely manner and ensure the effective supervision of decentralization.

Secondly, village clusters are more than just spatial or industrial divisions, they are a group of villages with common features or issues. To further grow these village groups, it is necessary to construct rural planning that targets rural group development and to determine the clusters’ development plans and investment priorities based on their features. The southern village clusters in Guiyang County are located in the county town area, which form a solid core area due to the concentration of enterprises, such as smart furniture manufacturing, high-tech industries, and technological agriculture. Their development mainly rely on the radiation of the main city, and they benefit from the reconstruction of rural industries. They can rely on horizontal-scale clustering to introduce more new material industries and high-end equipment manufacturing to improve the efficiency of traditional industries. The western village clusters are primarily located along the Chongling River, and they have a competitive advantage in the nonferrous metal industry, thus forming an important economic cluster. The areas should increase the diversity and complementarity of industrial projects, and improve resource efficiency within the group by optimizing capital, labor, and land resource allocation to constantly improve project competitiveness. The northern village clusters of Guiyang are dominated by agricultural projects, including production and processing projects. They are suitable to adopt the mode of “enterprises + cooperatives + cultivation bases” to develop the agricultural industry, create vertical scale clusters of upstream and downstream industrial mutual assistance, and rely on the advantages of agricultural product processing and planting bases. The northwest clusters have a comparative advantage in citrus and oil tea production, while the northeast clusters have a comparative advantage in tobacco production and processing. They are also local specialty agricultural products, which can be marketed through the network platform to boost sales. Furthermore, the village clusters in the central region are small in scale without a dominant development direction, and they are located in Guiyang’s ecological protection zone, which is subject to strict environmental restrictions, with green ecological agriculture and rural tourism potentially serving as the group’s breakthrough point.

China’s rural governance is gradually characterized by diversification, refinement, and innovation, emphasizing government’s leading role while encouraging the participation of all social forces. In contrast, Europe, the United States, Japan, and South Korea pay more attention to the self-governance capacity of rural communities and have formed a multi-level governance model through the participation of pluralistic subjects, which provides reference for China’s rural governance to shift to the grassroots [47]. Furthermore, the governance idea based on relational governance proposed in this study, particularly through policies based on decentralizing authority to rural communities, not only provides constructive ideas for rural development in China but also serves as a policy reference for developing countries, such as those in Asia and Africa, which are nearing the end of their urbanization process.

6.2. Conclusions

In the context of urbanization nearing its end and the transformation of China’s governance model, an in-depth discussion of the optimization of policy paths for the implementation of the rural revitalization strategy is not only a positive response to China’s current social development needs but also a forward-looking exploration of the future governance model of the countryside. Based on the relational governance perspective, this paper investigates how to use policy tools to encourage the establishment of an optimal rural cluster model. For this reason, this study selects typical county for analysis, employs the gravity model to measure the co-opetition relationships between villages, simulates village networks under assigned and meritocratic policy scenarios. Employing social network analysis, this paper analyzes network characteristics in various scenarios and compares the impact effects of different networks. Finally, it provides feedback for policy and rule formulation. The study’s findings demonstrate that village clusters developed under a meritocracy policy have a heavy reliance on resources and location, as well as severe pressure on the natural environment, a monolithic economic structure, and a weaker ability to resist risks. In contrast, the assignment strategy has significant advantages in exploring potential clusters and nurturing diverse clusters, and the resulting village cluster structure can better maintain the function of agricultural production, inherit local culture, and promote soil health. At the same time, this policy environment has the potential to stimulate the intrinsic vitality of the villages and attract more groups to participate in the construction of the village, thus assuring the long-term sustainability of more villages. As a result, this paper proposes that the county government decentralize power to villages and give them equal rights, thus allowing them to form diverse village clusters based on their comparative advantages.

This study has some limitations. Firstly, this study views competition and cooperation as an integrated relationship and did not distinguish them. Secondly, the number of simulation experiments is relatively limited in the course of the study, with only three groups of experiments conducted under each policy rule. To improve the study’s comprehensiveness and accuracy, the number of experimental groups should be increased for future in-depth exploration. Thirdly, this study treats a county as a complete unit for village governance, mitigating interference from adjacent villages in the network simulation conducted at the county level. Moving forward, the study will also investigate the differences in governance measures arising from variations between open boundaries and the internal governance networks of closed systems. Lastly, the Euclidean distance was used to determine the distance in the gravity model, consistent with the geographical principle that “geographically proximate entities often exhibit similarities.” No modifications were made to the distance calculations. In the next phase, the plan is to explore the differences in village governance measures between and within counties by considering variations in road distances, such as by using the Manhattan distance algorithm.