Research on the Effect and Mechanism of Provincial Construction Land Spatial Agglomeration Empowering Economic Resilience in China

Abstract

Exploring the effects and mechanisms of spatial agglomeration of construction land resources on economic resilience across Chinese provinces will provide theoretical support for governments to optimize the allocation of productive forces and enhance economic resilience through rational distribution of construction land quotas. Based on the “Structure-Conduct-Performance (SCP)” analytical framework, this paper identifies spatial agglomeration through the share of the largest city and draws on the microeconomic concept of “elasticity” that reflects the relationships between variables to construct economic resilience with spatial relationship attributes. On this basis, it utilizes China’s provincial panel data gathered since 2000 and employs fixed-effects models, mediation models, moderation models, quantile regression, and subsample regression to examine the impact mechanisms of the spatial agglomeration of construction land on economic resilience. The research finds the following: the spatial agglomeration of construction land has a positive empowering effect on economic resilience; innovation and technical efficiency are important transmission paths for the spatial agglomeration of construction land to empower economic resilience; and further research shows that the empowering effect has an inverted U-shaped process, with the promoting effect being predominant. The empowering effect increases with rising quantiles and exhibits regional heterogeneity, showing an ascending gradient from eastern to western regions. The basic law in the western region is consistent with that of the whole country, and the scale of provincial construction land will strengthen the empowering effect. The research findings can provide decision-making references for the implementation and deepening of the main functional area strategy, as well as for strengthening the concentrated allocation of construction land quotas to advantageous regions.

1. Introduction

Enhancing economic resilience and promoting sustainable economic growth is a major strategic task that China faces in the post-crisis era compounded by the COVID-19 pandemic. Resilience is a theoretical analytical tool for evaluating the shock-resistance ability of a subject. It was first applied in mechanics to reveal the ability of an object to return to its initial state after being subjected to pressure. Subsequently, this analytical tool gradually penetrated into other disciplines. In the field of regional economics, some scholars have pointed out that regional economic resilience is the ability of a regional economy to cope with external shocks [1], that is, the ability of a regional economy to effectively respond to external disturbances, resist risk shocks, and achieve sustainable economic growth when facing external shocks [2]. It is not difficult to see that regional economic resilience is a concept closely tied to “elasticity”, reflecting the adaptive capacity of economies within an open economic environment and revealing the relationship between regional economic systems and external economic conditions. Therefore, regional economic resilience is not only an important indicator for evaluating the ability of a regional economy to resist external threats and recover from an unfavorable environment, but also an analytical tool for evaluating its ability to obtain growth by utilizing external opportunities. An economy with strong resilience suffers less from negative shocks when facing external threats and can sustain rapid growth after the shocks subside; when facing external opportunities, it can obtain greater actual benefits from the opportunities and achieve more obvious and continuous growth. In summary, regional economic resilience reveals an indicator about the relationship between a regional economy and the external environment. This paper defines regional economic resilience as follows: against the background of the domestic cycle, the intensity of the response of a regional economy to the economic situation changes of other domestic economies. At the result level, it can be reflected by the relative economic growth level. Then, what is the internal driving mechanism of regional economic resilience?

On the one hand, agglomeration, as an important economic phenomenon and organizational mode of economic activities [3,4], provides an important perspective for explaining China’s economic resilience. Based on different objects, agglomeration can be identified through scale, intensity, and structure. In recent years, scholars have investigated the relationship between the agglomeration of different entities and economic resilience in the following ways: (1) From the perspective of industrial agglomeration patterns, most studies have examined the relationship between industrial diversification (including related and unrelated diversification) and urban economic resilience [5,6,7,8]. (2) From the perspective of specific industry, the direct impact and moderating effect of (digital) financial agglomeration were studied [9,10,11]. (3) From the perspective of population, the impact of population agglomeration on urban economic resilience was examined [12,13]. (4) From the perspective of economic spatial structure, the impact of urban polycentric spatial structure patterns on economic resilience was examined [14,15].

On the other hand, land resources serve as the material foundation for economic growth. Their contribution to economic growth is mainly reflected in three aspects: providing necessary construction land for national economic development, exerting a spatial carrying function, and acting as an important tool for local governments to strengthen macro-regulation, boost economic growth [16], and achieve the coordinated coupling of “population-economy” [17]. The existing literature has paid more attention to the construction land misallocation, and among them, some studies have found that the misallocation of construction land reduces labor productivity [18], leads to an increase in housing prices [19], hinders industrial structure upgrading [20], and triggers economic fluctuations [16]. A small number of studies have focused on the impact of land marketization and market-oriented transfer on economic resilience [21,22].

In summary, the existing literature has paid attention to the role of agglomeration in economic resilience, as well as the role of construction land resources in economic growth and economic resilience. However, few studies have examined the impact of spatial agglomeration in the spatial allocation process of construction land resources on economic resilience at China’s provincial level. There are two background facts about the allocation of construction land resources in China: (1) It follows a “top-down” and “region-first, use-second” quota allocation process. At the provincial level, affected by urban competition and other factors, a spatial agglomeration pattern of construction land scale distribution has been formed. (2) Since 2003, land management policies have focused on driving economic growth through the allocation of land resources. In 2004, the central government explicitly proposed enhancing the ability of land to participate in macroeconomic regulation. Construction land quotas have become an important means for the country to support economically backward regions. Following the central strategy of balancing regional development and the idea of using land to participate in macroeconomic regulation, policies have restricted the supply of construction land in large cities, leading to spatial mismatches in land supply and demand, which may in turn suppress the sustainable growth of spatial economy. Optimizing the spatial allocation of construction land resources, adjusting their spatial structure, and enhancing their spatial agglomeration is one of the important contents of construction land participation in macroeconomic regulation. From the perspective of provincial governments, construction land resources will become a key focal point. In the provincial territorial spatial planning, the allocation of construction land quotas among different cities is likely to be used as a policy tool to strengthen or alleviate the economic development gap within the province, and the limited construction land quotas are concentrated on cultivating growth poles to drive the overall economic development of the province [23]. From the perspective of the endogenous growth theory and structuralism, optimizing the allocation of productive forces at the provincial level and strengthening the spatial centralized allocation of scarce resources will establish a mechanism of increasing returns to scale through spatial agglomeration, mitigating the impact of external adverse conditions on provincial economic growth. Therefore, the spatial reallocation of construction land resources can provide new impetus for the next round of China’s economic growth [24]. Given the exogenously determined scale of construction land at the provincial level, based on the “Structure-Conduct-Performance (SCP)” logic paradigm [25,26], optimizing the internal spatial reallocation of construction land within provinces and enhancing the spatial agglomeration level of construction land resources may be an important means to improve the economic resilience of provinces.

Therefore, based on the “structural functionalism” methodology and the SCP analytical framework, this paper takes the spatial allocation of provincial construction land resources as the starting point to investigate the mechanism by which its spatial agglomeration empowers economic resilience, which provides decision-making references for the optimization of the spatial allocation of construction land resources and the formulation of spatial development strategies to enhance the economic resilience of Chinese provinces. The marginal contributions of this study are as follows: (1) Deepening and expanding the connotation and extension of resilience. The existing literature mostly employs a comprehensive evaluation method to measure the economic resilience of China’s regions based on dimensions such as “resistance” and “recovery.” This paper interprets regional economic resilience from the perspective of the relationship between the economy and the external environment, extending the “elasticity” concept based on variable relationships in microeconomics to the “spatial” relationship level in an organic manner to identify regional economic resilience. This approach makes up for the deficiency of the “spatial” aspect in the “elasticity” concept of microeconomics. (2) Expanding the application scenarios of the SCP analytical framework and structural functionalism. Organically extending the “structural functionalism” in sociology and the SCP analytical framework in the industrial organization theory of industrial economics to the field of resource spatial agglomeration (spatial structure) formed by the spatial allocation of resources in regional economics, and constructing an SCP analytical framework for the impact of resource spatial agglomeration on economic resilience, thus expanding the extension of the SCP model and proposing the conceptual model of the SER conduction path. (3) Enriching the research on agglomeration-driven mechanism of economic resilience. Agglomeration has an important impact on economic resilience, but the connotation and extension of agglomeration vary in different disciplines. This paper examines the driving mechanism of economic resilience from the perspective of resource allocation under resource scarcity and the spatial agglomeration of resource spatial allocation in regional economics, enriching the agglomeration-driven mechanism of economic resilience.

2. Policy Background, Analytical Framework, and Mechanism of Action

2.1. Policy Background

The spatial allocation of construction land is the specific implementation of the major function oriented zone strategy (MFOZS). The MFOZS is the guiding principle for optimizing the pattern of territorial space development, and optimizing the pattern of territorial space development is the goal and direction of the MFOZS and spatial allocation of construction land. Therefore, discussing the spatial allocation pattern of provincial-level construction land cannot be separated from the large framework of territorial space. The spatial allocation pattern of construction land is related to the planning policies for territorial space development. In 2007, the report of the 17th National Congress of the Communist Party of China first proposed to optimize the pattern of territorial development, strengthen territorial planning, improve regional policies, and adjust the economic layout in accordance with the requirements of forming main functional areas. Figure 1 summarizes the relevant planning policies since 2010.

Generally speaking, regarding the planning policies for territorial space development, on the one hand, they have shifted from territorial space to land use and from main functional areas to a certain type of functional area (urbanized areas); on the other hand, these planning policies have gone through two-stage evolution, and both stages emphasize “comparative advantage”, “complementary advantages”, “factor agglomeration”, and “major function oriented zone”, and ultimately, they all focus on the centralized spatial allocation of construction land quotas in advantageous areas. Specifically:

During the first stage (2010–2020), Document No.46 [2010] issued by the State Council put forward the basic concept of “major function oriented zone” in territorial space. On this basis, Document No.3 [2017] issued by the State Council proposed the concept of “agglomerated development” in territorial space, and the two concepts are consistent and successive. The economic logic behind the major function oriented zone is “specialization” and “comparative advantage”. Adhering to the principle of comparative advantage and carrying out specialized division of labor will guide the efficient spatial agglomeration of population and industries, then optimize the allocation of productive forces. The idea put forward at the Fifth Meeting of the Central Financial and Economic Commission of tilting construction land resources towards central cities and key urban agglomerations is a specific manifestation of the “major function oriented zone” and “agglomerated development”.

During the second stage (from 2021 to the present), the Outline of the 14th Five-Year Plan for National Economic and Social Development and the Long-Range Objectives Through the Year 2035 clearly proposes forming a new pattern of principal function zone and territorial space development and protection by leveraging “comparative advantages” and “factor agglomeration”. The 20th National Congress of the Communist Party of China (hereinafter referred to as the “20th CPC National Congress”) has further elevated the above ideas in the Outline, bringing the “major function oriented zone” to a new level. Guided by this, China will optimize the layout of major productive forces, ensure the rational flow and efficient agglomeration of various factors in space, and form a territorial space system featuring complementary strengths. The discussion on “land factor allocation” at the fourth meeting of the Central Commission for Deepening Reforms is an initiative to implement the guiding principles of the 20th CPC National Congress, aimed at optimizing the system for major functional zones and improving the national spatial development and protection pattern. In essence, without the spatially optimized allocation of land factors, there would be no formation of major functional zones or the establishment of an orderly and coordinated spatial development pattern. The discussion on “optimizing land management” at the Third Plenary Session of the 20th Central Committee of the Communist Party of China is in line with the ideas of the Fourth Meeting of the Central Commission for Comprehensively Deepening Reforms and the Fifth Meeting of the Central Financial and Economic Commission. All of them emphasize “land factor allocation”, which also represents an elevation of the issue of “spatial allocation of construction land resources”, as discussed at the Fifth Meeting of the Central Financial and Economic Commission. Essentially, these policies all stress the need to strengthen the supply of land resources in advantageous areas, that is, the spatial centralized allocation of construction land resources. For non-agricultural economic activities, advantageous areas refer to important agglomeration and carrying areas of population and economic activities within a certain geographical scope. The purpose of strengthening the allocation of construction land resources toward advantaged regions is to enhance spatial coordination and synergy between population and land elements in these areas, leverage the positive externalities generated by the concentration of various resources and factors within these regions, and thus form urban spaces with distinctive primary functions.

At the level of municipal territorial space, which is part of provincial territorial space, the changes in urban space directly affect the patterns of the other two types of space (agricultural and ecological). From the perspective of functional areas, construction land belongs to urbanized areas and bears urban functions. It is the most direct supporting factor for spatial development, and the spatial allocation pattern of construction land directly affects the territorial space pattern. At the provincial level, comprising different prefecture-level cities, the allocation of urban construction land quotas is centrally managed by provincial governments. The varying major functional zones of different cities determine spatial disparities in construction land resource allocation levels, thereby forming a spatial allocation framework for provincial construction land resources.

2.2. Analytical Framework

As mentioned above, China’s provincial-level spatial allocation of construction land refers to the differentiated distribution of urban space among different cities under the major function-oriented zone system, exhibiting distinct systematic structural characteristics. From the perspective of structural functionalism, this paper extends the “Structure-Conduct-Performance” analytical framework in the field of industrial organization to the field of regional spatial organization, and then obtains the analytical framework shown in Figure 2. That is, the spatial structure affects the spatial performance through spatial behavior, and to a certain extent, the spatial performance in turn requires the adjustment of the spatial structure to generate positive spatial external behavior.

2.2.1. Spatial Agglomeration of Spatial Allocation of Construction Land Resources

Optimizing the spatial allocation of resources to achieve maximum overall economic efficiency is one of the core issues in regional economic research. The spatial allocation of construction land resources is based on the optimization of the spatial structure of construction land resource elements led by the government, and is the sum of the allocation and flow of construction land in different geographical units in space. It has the following connotations and characteristics.

First, construction land resources are scarce. As mentioned above, the territorial space is a spatial functional division system formed by the mutual influence and restriction of the three major spaces: urban, agricultural, and ecological. Whether it is the land use control system focusing on cultivated land protection and ensuring food security in the past, or the territorial space use control system that combines the protection of agricultural and ecological spaces in the current stage of ecological civilization construction, due to the multiple functional attributes of land resources, construction land resources have obvious scarcity.

Second, the government intervenes in the allocation of construction land resources. On the one hand, unlike other resources, construction land does not directly act on economic activities; however, it is the carrier of non-agricultural economic activities and plays a leading role in the process of economic and social development. Due to the scarcity of the land, the government needs to guide economic and social development through the allocation of construction land quotas; on the other hand, the allocation of construction land resources is a process of non-agriculturalization of agricultural land and unused land, and its allocation is related to food security and ecological security, so the government needs to conduct macro-control.

Third, the provincial-level allocation of construction land quota has spatial characteristics. Economic activities are inseparable from space, and construction land resources need to be allocated to different geographical units in space. In terms of the allocation method, it is mainly manifested as a “top-down” planned means of decomposition layer by layer from the central government to provincial governments and then to prefecture-level cities. After the total construction land quota is allocated to provincial governments, the optimal allocation of construction land resources among cities reveals the spatial allocation characteristics of construction land resources.

Fourth, the spatial allocation of provincial-level construction land exhibits agglomeration characteristics. Due to the spatial heterogeneity of endowments and the inherent incomplete divisibility of economic activities, not all geographical units within a province can offer equal opportunities for productive activities and spatial development. This compels construction land resources to flow towards geographical units with higher marginal returns, thereby leading to the spatial agglomeration of construction land allocation. At the implementation level of spatial allocation, provincial governments conduct overall planning and macro-control based on the development conditions of each city. On the premise of ensuring the maximization of the province’s overall interests, they balance equity and efficiency, selectively encouraging or restricting the development of certain cities. Ultimately, through the combination of “planning,” “plan,” and “approval and licensing,” construction land resources are allocated, resulting in a significant spatial agglomeration of construction land resources (Figure 3).

2.2.2. The Revealing of Regional Economic Resilience on Regional Spatial Relations

Regional economic resilience is a derivative of the ‘elasticity’ theory in Microeconomics, reflecting spatial dependence and response dynamics. On the one hand, regional economic resilience falls within the research scope of regional economics and is related to the issue of the sustainability of regional economic growth. There exists a close spatial relationship among regions, and regional economic resilience implies spatial relationships. On the other hand, regional economic resilience follows the basic logic of general economics and is a concept related to “elasticity”. Elasticity has the following two important characteristics: ① The essence of elasticity is to reveal relationships. ② Existing elasticity concepts pay more attention to “variable relationships”, for example, some scholars have used this method to reveal the relationship between population and land [17] and neglect “spatial relationships”, especially spatial economic relationships. In fact, the economic growth of different regions (spaces) is not isolated but shows obvious spatial correlation. Thus, we obtain the regional economic resilience (elasticity) that reflects “spatial relationships”, that is, the degree of intensity with which a regional economy responds to changes in the economic situation (external environment) of other domestic regional economies, or the ability of a regional economy to cope with changes in the external environment. This can be expressed as the economic resilience R of region i in year t: R = (ΔG_i/G_i)/(ΔG/G), where G_i represents the GDP of region i and ΔG_i represents the period change in the GDP of region i, so G_i/ΔG_i represents the GDP growth rate of region i; G represents the national GDP excluding region i, ΔG represents the change in the national GDP excluding region i, and similarly, G/ΔG represents the GDP growth rate of the nation excluding region i. This indicator effectively integrates the “elasticity” concept that reflects variable relationships in microeconomics with a “spatial” perspective, thereby revealing spatial relationships.

2.2.3. An Analysis Framework of “Structure-Conduct-Performance” for the Economic Resilience Empowered by the Spatial Agglomeration of Construction Land Resources

Structuralism is an important methodology. Whether it is the structural functionalism in sociology or the “Structure-Conduct-Performance (SCP)” analytical paradigm in industrial organization theory, both emphasize the importance of “structure”. In terms of regional spatial structure, it reveals the spatial distribution state and combination form of various economic activities and their elements within a certain region. It is a spatial organization formed by the interconnection and interaction of various economic entities in the region, and is characterized by the form and degree of spatial agglomeration. It plays an important role in the process of regional economic development. The spatial agglomeration of resources reveals the non-equilibrium spatial distribution characteristics of resources and reflects the spatial structure of resources. Based on the structural functionalist methodology, which emphasizes structural functionality and holism, different spatial structures reflect distinct spatial production relations, thereby generating differentiated effects in terms of sharing, matching, and learning, ultimately leading to varied spatial performance.

Among them, the sharing mechanism helps to apportion fixed costs and enhance economic efficiency; the matching mechanism facilitates the connection of supply and demand and improves resource allocation; the learning mechanism generates knowledge spillovers and innovation diffusion, forming an innovation network and creating human capital externalities. From the perspective of the SCP analytical paradigm, under the constraint of the scarce total amount of provincial construction land resources, optimizing its spatial allocation and enhancing spatial concentration will guide the spatial agglomeration of other production factors dependent on construction land; through the effects of spatial agglomeration externalities, this will significantly impact regional economic resilience. From this, and referring to the approach in this literature [27], we can find the SER conduction path model as shown in formula (1). Where R is the regional economic resilience, S is the spatial agglomeration of construction land, and E is the externality of spatial agglomeration. F(S) reveals the conduction path function of S affecting R.

$$R=S\times F(S)=S\times {\displaystyle \frac{E}{S}}\times {\displaystyle \frac{R}{E}}$$

(1)

2.3. Mechanism of Action

Based on the SCP analytical framework and the SER conceptual model, this paper further constructs the mechanism of spatial externalities in the process of empowering economic resilience through the spatial agglomeration of construction land (Figure 4).

A reasonable and orderly spatial organizational structure is an important source for improving the quality and efficiency of economic growth [28], and it is also an important guarantee for coping with the “crises” and “opportunities” in the economy and enhancing economic resilience. From the perspective of land use, the provincial-level land resource allocation refers to a spatial division system of functional structures formed through the allocation of construction land among cities at different levels, guided by principles such as major function-oriented zones. The spatial agglomeration of construction land refers to the non-equilibrium or unbalanced allocation of construction land resources among different cities, which reveals the spatial relationship among provincial-level cities. In essence, it is an organizational network with a certain hierarchy established with the spatial allocation of construction land resources as the driving force and the goal of achieving the common development of all cities in the region. It affects the sustainable development of the entire province through the collaborative interaction of nodal cities.

Specifically, first, the spatial agglomeration of construction land will guide various micro-economic entities (especially low-productivity enterprises) to concentrate in central cities with an urbanized economic form. This will help give full play to the comparative advantages of provincial-level central cities in terms of knowledge, technology, information, and talent, enabling those low-productivity micro-economic entities to “huddle together for warmth” in the face of “crises” and “opportunities”, thus forming an agglomeration effect of economic resilience. Second, the spatial agglomeration of construction land guides the spatial agglomeration of economic activities, which will produce a strong price index effect, reducing the production costs of enterprises. At the same time, under the effect of the spatial competition, medium- and low-labor-productivity enterprises are forced to continuously improve their production efficiency, thus having a positive impact on the economic resilience of the entire region.

Hypothesis 1:

The spatial agglomeration of construction land helps to enhance economic resilience.

The spatial agglomeration of construction land affects economic resilience through innovation. Regarding the relationship between the spatial agglomeration of construction land and innovation, the spatial agglomeration of construction land provides conditions for the geographical concentration of innovation subjects, thereby enhancing the level of knowledge spillover and improving the regional innovation ability. Specifically, first, the spatial agglomeration of construction land helps to form an inclusive and diversified innovation development environment. The spatial agglomeration of construction land drives people with different skill levels, cultural backgrounds, and ideas to communicate with each other on the basis of geographical concentration, promoting the spatial spread of coded and non-coded knowledge in scientific and technological innovation activities and giving rise to innovation and creativity. Second, the spatial agglomeration of construction land helps to form a stable relationship network that promotes the spillover of innovative knowledge. The spatial agglomeration of construction land makes industries or enterprises engaged in related and similar activities consciously or unconsciously cluster together, creating innovation growth poles. This helps the rapid spread of various information, knowledge, and “best practices” across the region, driving “collective learning” among enterprises in a specific area and the competition and imitation among entrepreneurs, thus promoting the birth of new industries and new enterprises. Third, the spatial agglomeration of construction land helps to form a super-large-scale market that stimulates innovation. The spatial agglomeration of construction land means the coordinated co-agglomeration of population and industrial economic activities on land elements, which will give rise to large cities and build large markets. In the process of meeting the demand for diverse and differentiated products, new ideas and new product R & D are stimulated, thereby improving the overall regional innovation ability. Fourth, the spatial agglomeration of construction land helps to drive the agglomeration of new-quality productive forces for innovation. Based on the spatial classification and spatial selection theory [29], the spatial agglomeration of construction land resources will promote the coordinated spatial agglomeration of high-endowment enterprises or labor and land resources, improve the efficiency of resource allocation, reduce the land-use cost for high-endowment enterprises and labor elements to agglomerate in central cities, weaken the congestion cost in the process of spatial agglomeration of high-quality production factors, and thus improve the innovation ability.

Regarding the relationship between innovation and economic resilience, innovation, as a key adaptive factor, is the most fundamental driving force for economic and social structural transformation and endogenous economic growth [30]. It reveals a region’s learning capacity and the level of knowledge diversification. Learning capacity makes a region more effective in adaptation during and recovery from major shocks [31]. Places identified as innovation leaders during a crisis are more likely to withstand the crisis or recover quickly from it [32]. A region with diversified knowledge capacity has a stronger ability for the birth of new enterprises or the revival of bankrupt enterprises [33]. In other words, innovation is the primary force driving sustainable economic growth. It can bring new industries and new business forms to a region, enabling the regional economy to efficiently utilize and allocate existing knowledge resources after a shock, make adaptive structural adjustments, and seek potential development opportunities. This injects important momentum into the recovery and sustained growth of the economy after a shock, thereby enhancing economic resilience. Specifically, first, innovation will drive the development of emerging industries such as artificial intelligence, big data, and cloud computing, and it also drives the emergence of new business forms such as the sharing economy, platform economy, and digital economy, becoming new economic growth points. Second, innovation promotes the evolution of industries from low-value-added links to high -value-added links, realizes the transformation of old and new driving forces, enhances regional competitive advantages, and strengthens the economy’s ability to resist shocks. Third, innovation stimulates investment, gives rise to new products, and stimulates new consumption, enhancing economic resilience. The transformation of new technologies corresponding to new concepts into new products requires a large amount of investment, and investment will produce a multiplier effect, stimulating economic growth and enhancing economic resilience. At the same time, the launch of new products brought about by innovation investment will stimulate new consumer demand, which in turn drives economic growth and enhances economic resilience.

Hypothesis 2:

The spatial agglomeration of construction land resources empowers economic resilience through innovation.

The spatial agglomeration of construction land resources affects economic resilience through technical efficiency. From the perspective of output, technical efficiency is the ratio of the actual output of a production unit to the ideal maximum possible output under the same input. A higher technical efficiency means that limited resources can be utilized to the maximum extent to produce or provide more and better products or services, which means less inefficiency. The spatial agglomeration of construction land reveals the trend of the concentration of scarce construction land resources in advantageous areas. At the micro-level, on the one hand, it will guide individual enterprises to make centralized location selections in advantageous areas, which helps to improve the sharing level of resource services, increase the matching speed of labor, and reduce the transaction costs and the costs of knowledge, technology, and information dissemination among enterprises; on the other hand, it will accelerate the flow of labor to advantageous areas, forcing the spatial agglomeration of labor resources and improving the spatially decentralized allocation of labor resources, achieving a Pareto improvement in the spatial allocation of labor, and thus improving technical efficiency and enhancing economic resilience. That is, within a specific region, for the same quantity of labor resources, their spatial concentration optimizes the spatial distribution structure of labor resources, provides a labor pool for the economic development of advantageous areas, reduces the spatial misalignment of labor in the production process, and then improves technical efficiency through the spatial structure effect. At the meso-level, the centralized location selection of enterprises will prompt enterprises in different industries and different enterprises in the same industry to form an industrial cluster organizational pattern with a co-existence of diversified and specialized agglomeration in space. This industrial geographical agglomeration is conducive to the expansion of the intermediate input market scale, thus providing specialized and low-cost intermediate inputs for upstream and downstream enterprises. This not only reduces the production costs of enterprises in the industry but also improves the industry efficiency loss caused by the spatial segmentation of the industry product value chain. At the macro-level, it will adjust the spatial structure of economic activities, realize the spatial reorganization of economic activities, and promote the spatial concentration of economic activities. On the one hand, it reduces the production costs, including transportation costs and the economic efficiency loss caused by information asymmetry and information dissemination lag brought about by the decentralized layout of economic activities, thereby enhancing economic resilience.

Hypothesis 3:

The spatial agglomeration of construction land resources empowers economic resilience through technical efficiency.

3. Research Design

3.1. Model Specification

3.1.1. Fixed-Effects Model

Model (2) examines the overall impact of the spatial agglomeration of construction land resources on economic resilience across China’s provincial-level regions.

$$R_{it}={\alpha }_0+{\alpha }_1{\mathrm{S}}_{it}+{\displaystyle \sum _{k=1}^m{\varphi }_k}{X}_{kit}+{\mu }_i+{\tau }_t+{\epsilon }_{it}$$

(2)

where the subscripts i and t represent provincial region i and year t, respectively; R is economic resilience; S is the spatial agglomeration of construction land resources; X is the control variable; m represents the number of control variables; μ_i and τ_t are individual and time fixed effects; and ε_it is the random disturbance term.

3.1.2. Four-Stage Mediating Effect Model

Based on the conceptual model of the SER conduction path and the research hypothesis, this paper tests the conduction path through which the spatial agglomeration of construction land empowers economic resilience. This paper employs a four-step mediation effect testing method. Models (3) to (5) are established to examine the mediating mechanism of the spatial agglomeration of construction land in enhancing economic resilience at the provincial level. Sobel and Bootstrap tests are conducted to enhance the completeness and credibility of the mechanism testing. In the equations, E represents the mediating variable, namely agglomeration externalities, which specifically include innovation (I) and technical efficiency (T). The symbols for other variables are the same as above.

Equation (3) is the determinant model of the mediating variables affected by the spatial agglomeration of construction land. If the relationship is significant, proceed to the next step. Equation (4) is the relationship model of the impact of the mediating variables on economic resilience. If the relationship is significant, proceed to the next step. Equation (5) is the determinant model of economic resilience that incorporates both the spatial agglomeration of construction land and the mediating variables. In the empirical analysis, observe the changes before and after the impact of the spatial agglomeration of construction land (S) on economic resilience (R). According to the principle of the mediating effect model, if the coefficients b₁, β₁, and c₂ are all significant, the sign of the product of b₁ and c₂ is the same as that of c₁, and the coefficient c₁ becomes smaller or less significant compared with α₁, which indicates the existence of a mediating effect.

$$E_{it}=b_0+b_1{S}_{it}+{\displaystyle \sum _{k=1}^{\mathrm{m}}{\theta }_k}{X}_{kit}+{\mu }_i+{\tau }_t+{\epsilon }_{it}$$

(3)

$$R_{it}={\beta }_0+{\beta }_1{E}_{it}+{\displaystyle \sum _{k=1}^m{\varphi }_k}{X}_{kit}+{\mu }_i+{\tau }_t+{\epsilon }_{it}$$

(4)

$$R_{it}=c_0+c_1{S}_{it}+c_2{E}_{it}+{\displaystyle \sum _{k=1}^m{\varphi }_k}{X}_{kit}+{\mu }_i+{\tau }_t+{\epsilon }_{it}$$

(5)

3.2. Variable Measurement

3.2.1. Explained Variable

Provincial economic resilience (R) refers to the intensity of the response of a provincial economy to the changes in the economic situation (external environment) of other domestic economies. As previously mentioned, the existing literature mostly measures the economic resilience of Chinese provinces based on dimensions such as “resistance” and “recovery” using a comprehensive evaluation method. And different scholars use different indicator systems. As a result, the comprehensive evaluation method not only lacks a universally recognized and reasonable indicator system and weights, but also has the drawback that indicators can easily confuse cause and effect. In recent years, some research has begun to identify economic resilience based on the difference in economic growth rates. Among them, some studies characterize provincial economic resilience based on the difference between provincial growth rates and the national economic growth rate [34,35]. The more the literature indicates provincial economic resilience based on the ratio of this difference to the absolute value of the national economic growth rate. But the economic significance of the latter part of the literature is not clear, or rather, it lacks a very clear economic theory to support it. This paper follows the “elasticity” theory in microeconomics. Considering that the sign of the changes in the economic situation of other economies is uncertain over time, based on this, this paper uses Equation (6) to identify provincial economic resilience. In the equation: R_A_it is the economic resilience of province i in year t. If the value of R_A_it is larger, it indicates that, compared with the national average level, the economy of province i in year t has a stronger ability to resist the adverse effects of shocks or a stronger ability to utilize external opportunities for growth. Among them, ΔG_it = G_it − G_it−1 and G_it and G_it−1 are the real GDP of province i in year t and year t − 1 based on the base period of 1999. Therefore, ΔG_it/G_it−1 represents the GDP growth rate of province i in year t. Similarly, ΔG_t = G_t − G_t−1, where G_t =∑G_jt, G_t−1 = ∑G_jt−1, and j ≠ i, that is, G_t and G_t−1 represent the sum of the real GDP of other provinces except province i in year t and year t − 1. Therefore, ΔG_t/G_t−1 represents the national GDP growth rate excluding province i in year t.

$$R\_A_{it}={\displaystyle \frac{\Delta G_{it}}{G_{it-1}}}-{\displaystyle \frac{\Delta G_t}{G_{t-1}}}$$

(6)

3.2.2. Explanatory Variable

Provincial construction land resource spatial agglomeration (S) refers to the degree of imbalance in the spatial allocation of provincial construction land resources. The methods commonly used to measure its spatial agglomeration include non-parametric estimations, such as the share of the largest city and Herfindahl Index, as well as parametric estimations, like the Zipf Index. The larger the value, the higher the degree of provincial construction land spatial agglomeration. Comparatively speaking, the former can more accurately identify the degree of spatial agglomeration. Based on this, this paper uses the share of the largest city. The calculation method is shown in formula (7), where Area_irt represents the built-up area of the prefecture-level city with the r-th rank in the scale of urban construction land in province i in year t, and n represents the number of prefecture-level cities in province i. This indicator shows that the larger the share of the construction land scale of the largest city (r = 1) in the total construction land scale of all cities in the province, the higher the degree of provincial construction land resource spatial agglomeration.

$$S={\displaystyle \frac{Are{a}_{i1t}}{{\displaystyle \sum _{r=1}^{n}Are{a}_{irt}}}}$$

(7)

3.2.3. Mediating Variables

Innovation (I) is generally measured using patent applications or patent grants. This paper mainly uses the logarithm of patent application grants to identify and reflect the innovation level of a region. Technical efficiency (T) is measured using methods such as Stochastic Frontier Analysis (SFA) and Data Envelopment Analysis (DEA). Drawing on the research methods of the relevant literature [36], this paper selects the SFA of the trans-log production function to measure technical efficiency. The specific calculation methods are as follows.

$$\begin{array}{c}\ln Y_{it}=f(\ln K_{it},\ln L_{it},t)+\left({\nu }_{it}-u_{it}\right)={\beta }_0+{\beta }_1\ln K_{it}+{\beta }_2\ln L_{it}\\ {\beta }_{3}t+{\displaystyle \frac{1}{2}}{\beta }_4{\ln }^2{K}_{it}+{\displaystyle \frac{1}{2}}{\beta }_5{\ln }^2{L}_{it}+{\beta }_6\ln K_{it}\ln L_{it}+\\ {\displaystyle \frac{1}{2}}{\beta }_7{t}^2+{\beta }_{8}t\ln K_{it}+{\beta }_{9}t\ln L_{it}+\left({\nu }_{it}-u_{it}\right)\end{array}$$

(8)

In the formula, Y_it represents the actual GDP, and f (lnK_it, lnL_it, t) represents the frontier GDP output under full efficiency. K_it, L_it, and t represent capital, labor, and the time trend term, respectively. v_it−u_it is the composite error term, where v_it and u_it are independent of each other. v_it is a general stochastic disturbance term, which satisfies ${\nu }_{it}\sim N(0,{\sigma }_v^2)$. u_it is the technical inefficiency term, satisfying $u_{it}\sim N^{+}(u,{\sigma }_u^2)$. The technical efficiency (TE) for province i in year t can be calculated using Equation (9).

TE_it = exp(−u_it)

(9)

3.2.4. Control Variables

Scale of construction land (scale) is characterized by the logarithm of the built-up area of provincial-level cities. Level of foreign direct investment (fdi) is expressed as the proportion of foreign direct investment converted into RMB in GDP. Level of human capital (edu) is expressed as the logarithm of the weighted proportion of the education levels of the population over 6 years old, with the years of education as the weight. Level of industrial diversification (div) is characterized by the relative diversification index of the manufacturing industry. Intensity of environmental regulation (eg) reveals the motivation for transformation and innovation related to economic resilience, and is expressed as the share of investment in industrial pollution control in the total industrial output value. Degree of government intervention (gov) is characterized by the ratio of general budget expenditure to GDP. Upgrading of industrial structure (ais) is characterized by the ratio of the added value of the tertiary industry to that of the secondary industry. Level of urbanization (urb) is expressed as the proportion of the permanent urban population in the total population. Level of financial development (fin) is expressed as the proportion of the sum of deposits and loans in GDP. Regarding manufacturing agglomeration (agg), the location quotient is used to measure the level of industrial agglomeration, that is, the ratio of the share of manufacturing in local industry to that in national industry. Level of transportation infrastructure (railway) is characterized by the ratio of railway mileage to the administrative area.

3.3. Sample Selection and Variable Statistics

In terms of sample selection for the study, since the measurement of the spatial agglomeration of provincial-level construction land resources is based on a certain number of urban units, four municipalities directly under the Central Government of Beijing, Tianjin, Shanghai, and Chongqing, two special administrative regions of Hong Kong and Macao, and provinces (autonomous regions) such as Xinjiang, Tibet, Qinghai, Hainan, and Taiwan with a small number of cities or a large amount of missing data are omitted. Finally, the spatial units in this study involve 23 provincial-level regions in mainland China. In terms of time series, the time span is from 2000 to 2022. There are a total of 529 sample observations (23 per year × 23 years). In addition, the original data of the variables in this paper come from the annual “China Urban Construction Statistical Yearbook”, “China Statistical Yearbook”, “China Industrial Statistical Yearbook”, “China Environmental Statistical Yearbook” and the statistical yearbooks of each province or autonomous region.

Table 1. Basic statistics of variables.

Variable Properties	Variable	Variable Symbol	Mean Value	Standard Error	Minimum Value	Maximum Value
Explained variable	Economic resilience	R_A	0.011	0.020	−0.098	0.084
Explanatory variable	Construction land spatial agglomeration	S	0.320	0.125	0.134	0.721
Mediating variables	Innovation	I	9.363	1.742	5.366	13.679
	Technical efficiency	T	0.377	0.229	0.043	0.978
Control variables	Scale of construction land	scale	6.926	0.702	4.678	8.676
	Foreign direct investment level	fdi	0.022	0.019	0.000	0.105
	Level of industrial diversification	div	2.409	1.076	0.417	9.754
	Level of financial development	fin	2.728	0.740	1.413	4.969
	Manufacturing agglomeration	agg	0.982	0.143	0.588	1.501
	Upgrading of industrial structure	ais	1.018	0.258	0.518	1.953
	Degree of government intervention	gov	0.204	0.085	0.069	0.465
	Level of human capital	edu	0.016	0.007	0.002	0.034
	Intensity of environmental regulation	eg	0.004	0.004	0.000	0.029
	Level of urbanization	urb	0.499	0.124	0.233	0.748
	Level of transportation infrastructure	railway	0.020	0.019	0.004	0.391

reports the basic statistics of the variables.

4. Empirical Analysis

4.1. Benchmark Regression

Table 2. Benchmark Regression.

Variables		Fixed Effect			Random Effect
	(1) R_A	(2) R_A	(3) R_A	(4) R_A	(5) R_A	(6) R_A
S	0.038 **	0.052 ***	0.046 **	0.015	0.034 **	0.043 ***
	(0.02)	(0.02)	(0.02)	(0.01)	(0.01)	(0.01)
scale		0.019 ***	0.015 *		0.008 **	0.009 **
		(0.01)	(0.01)		(0.00)	(0.00)
fdi			0.151 ***			0.166 ***
			(0.06)			(0.05)
div			0.001			0.001
			(0.00)			(0.00)
agg			0.013			0.007
			(0.01)			(0.01)
is			0.002			−0.002
			(0.01)			(0.01)
gov			0.180 ***			0.139 ***
			(0.04)			(0.03)
edu			−0.439			−0.576
			(0.52)			(0.39)
hjgz			−0.277			−0.304
			(0.28)			(0.27)
fin			−0.013 ***			−0.012 ***
			(0.00)			(0.00)
urb			0.091 **			0.036 *
			(0.04)			(0.02)
railway			0.053			0.035
			(0.04)			(0.04)
Constant	−0.002	−0.126 ***	−0.144 ***	0.005	−0.052 **	−0.067 **
	(0.01)	(0.04)	(0.06)	(0.01)	(0.02)	(0.03)
Time effect	Yes	Yes	Yes	Yes	Yes	Yes
N	529	529	529	529	529	529
R2	0.291	0.302	0.398	0.289	0.298	0.389

Note: The standard errors are in parentheses, with *, **, and *** representing significant results at the 0.10, 0.05, and 0.01 levels, respectively. The same applies below.

reports the regression estimation results of fixed-effect and random-effect models regarding the impact of the spatial agglomeration of construction land resources on economic resilience. Through the Hausman test, this paper mainly conducts analysis based on the fixed-effect model. The report shows that, regardless of whether control variables are considered or not, the spatial agglomeration of construction land resources is significantly positive at least at the 5% level, indicating that the spatial agglomeration of construction land resources contributes to enhancing economic resilience. This suggests that, in the context of scarce construction land resources within the province, optimizing the spatial structure of provincial internal construction land resources, moderately strengthening the allocation of construction land resources towards central cities, and enhancing their spatial agglomeration is beneficial for improving the overall economic resilience of the province. Therefore, Hypothesis 1 holds. The reason is that the spatial agglomeration of construction land helps to improve the carrying capacity of central cities as advantageous regions and enhances their ability to cope with changes in the external environment. From the perspective of structural functionalism, the unequal allocation relationship of construction land resources between central and non-central cities revealed by the spatial agglomeration of construction land restructures the spatial organization of resource elements for provincial development, forms a corresponding spatial organization order, and endows the function of enhancing the overall economic resilience of the province.

4.2. Robustness Test

This paper verifies the robustness of the conclusions by replacing the explained variable. Three methods are further adopted to identify economic resilience as follows.

First, the economic resilience is measured by the economic growth rate in logarithmic form. Considering that the GDP growth rate of province i in year t, ΔG_it/G_it, is approximately equal to the growth rate in logarithmic form, ln(G_it/G_it−1) = lnG_it − lnG_it−1 =ln[(G_it−1 + ΔG_it)/G_it−1)] = ln(1 + ΔG_it/G_it−1), the method for measuring the economic resilience of province i in year t as shown in Equation (10) is obtained. That is, by comparing the actual GDP growth rate of province i with the actual GDP growth rate of the whole country excluding province i, an indicator for measuring the relative change in GDP is obtained. This indicator can also clearly reveal the intensity of the response of the provincial economy to the economic situation changes of other domestic economies and its relative economic growth ability, and can well reflect the idea of “elasticity”.

$$R\_B_{it}=\left(\ln G_{it}-\ln G_{it-1}\right)-\left(\ln {\displaystyle \sum _{j\ne i}^N{G}_{jt}}-\ln {\displaystyle \sum _{j\ne i}^N{G}_{jt-1}}\right)$$

(10)

Second, taking the year of the 2008 economic crisis as the benchmark, the actual GDP growth rate of each year in the province is compared with that in 2008, and an indicator for measuring the relative change as shown in Equation (11) is established. In the equation, the larger the value of R_C_it, the stronger the economic resilience; conversely, the weaker the economic resilience.

$$R\_C_{it}=\left(\ln G_{it}-\ln G_{it-1}\right)-\left(\ln G_{i2008}-\ln G_{i2007}\right)$$

(11)

Third, for comparative research, this paper further refers to the approach in the existing literature [34,35] and identifies economic resilience based on the national economic growth rate, as specifically shown in Equation (12). The difference between this equation and Equation (10) is that the observed province i is not excluded from the national economy as the reference object.

$$R\_D_{it}=\left(\ln G_{it}-\ln G_{it-1}\right)-\left(\ln G_{\mathrm{t}}-\ln G_{t-1}\right)$$

(12)

Columns (1)–Columns (3) of

Table 3. Robustness test.

Variable	(1) R_B	(2) R_C	(3) R_D	(4) R_A	(5) R_B	(6) R_C	(7) R_D
S	0.039 **	0.037 **	0.037 **	0.066 ***	0.059 ***	0.056 ***	0.056 ***
	(0.02)	(0.02)	(0.02)	(0.02)	(0.02)	(0.02)	(0.02)
Control variable	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Time effect	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Constant	−0.136 ***	−0.158 ***	−0.125 **	−0.193 ***	−0.176 ***	−0.198 ***	−0.168 ***
	(0.05)	(0.05)	(0.05)	(0.06)	(0.05)	(0.05)	(0.05)
R2	0.395	0.837	0.392	0.438	0.440	0.781	0.438

Note: The standard errors are in parentheses, with **, and *** representing significant results at the 0.05, and 0.01 levels, respectively. Due to layout limitations, estimates of control variables are omitted; the same applies below.

report the relationship between economic resilience and the spatial agglomeration of construction land under different measurement methods. The results show that the spatial agglomeration of construction land is significantly positive at least at the 5% level. By comparison, when the economic resilience is R_B and R_D, the impact intensity of the spatial agglomeration of construction land on the latter is relatively smaller than that on the former. In conclusion, regardless of how the explained variable is changed, the spatial agglomeration of construction land resources has a significant promoting effect on economic resilience. Therefore, the conclusion of the benchmark regression is robust.

On the other hand, considering the impact of the pandemic on China’s economy since 2020, this paper conducts robustness tests using data from 2002 to 2019. The results, as shown in columns (4) to (7) of

Table 4. Instrumental variable estimation.

Variable	The First Stage	The Second Stage
	(1) S	(2) R_A	(3) R_B	(4) R_C	(5) R_D
IV (exchange rate/dialect diversity)	0.046 *** (0.01)
S		0.219 ** (0.10)	0.191 ** (0.09)	0.188 ** (0.09)	0.188 ** (0.09)
Control variable	YES	YES	YES	YES	YES
Time effect	YES	YES	YES	YES	YES
R2	0.341	0.294	0.298	0.809	0.290
Cragg–Donald Wald F	19.724

Note: The standard errors are in parentheses, with **, and *** representing significant results at the 0.05, and 0.01 levels, respectively.

, indicate that regardless of the method used to identify provincial economic resilience, the spatial agglomeration of construction land resources remains statistically significant at the 1% level. This suggests that the pandemic has, to some extent, diluted the promoting effect of land use spatial agglomeration on economic resilience.

In summary, it can be seen that whether replacing the measurement method of the explained variable, provincial economic resilience, or adjusting the sample size, both indicate that the spatial agglomeration of construction land significantly promotes economic resilience. Moreover, after removing the samples affected by the pandemic shock, the promoting effect of the spatial agglomeration of construction land on economic resilience is significantly enhanced.

4.3. Endogeneity Discussion

This study has the following two potential endogeneity problems. First, although the spatial agglomeration of construction land is mainly the result of overall arrangements made by provincial governments based on the development of each city, it cannot be completely ruled out that provincial governments may tilt construction land resources towards central cities to build provincial economic growth poles and improve the overall economic resilience of the province, which may lead to a reverse causality between the spatial agglomeration of construction land and provincial economic resilience. Second, in the process of model setting, due to factors such as unobservable variables, there may be a problem of omitting important variables.

This paper constructs instrumental variables for the spatial agglomeration of construction land based on dialect diversity and exchange rates. First, drawing on the research results of Ding, C. et al. (2020) [37], dialect diversity is used as an instrumental variable for the spatial agglomeration of provincial construction land. The reason is that dialect diversity will, to a certain extent, lead to administrative division and thus affect the spatial agglomeration of construction land, meeting the relevance assumption. As a result of factors such as small-scale peasant economy, social fragmentation, population migration, and geographical barriers, dialects are an objective historical existence. And the process of dialect formation and development happened long ago, so it can be considered to have little correlation with the current regional economic resilience, meeting the exogeneity assumption. Second, since dialect diversity is cross-sectional data that does not change over time, to construct panel data, drawing on the research results of Davis, D.R. et al. (2002) [38], the exchange rate, a macro exogenous shock, is used as an instrumental variable, and the exchange rate is positively correlated with the agglomeration of economic activities. On this basis, the exchange rate is multiplied by the reciprocal of dialect diversity to obtain the instrumental variable for the spatial agglomeration of construction land.

The estimation results reported in Table 4 show that the Cragg–Donald Wald F statistic is 19.724, significantly greater than 10, indicating that there is no problem of weak instrumental variables. The first-stage regression estimation results show that the instrumental variable (IV) is significantly and positively correlated with the spatial agglomeration of construction land resources at the 1% level. The second-stage regression estimation results show that, regardless of the method used to identify economic resilience, the spatial agglomeration of construction land has a significant positive impact on economic resilience at the 5% level. Compared with the benchmark regression, the significance remains unchanged. This indicates that after overcoming the endogeneity problem, the conclusion of the benchmark regression still holds.

4.4. Mechanism Test

Table 5. Test of the transmission mechanism.

Variable	(1) R_A	(2) I	(3) R_A	(4) R_A	(5) T	(6) R_A	(7) R_A
S	0.046 **	1.029 ***		0.040 **	−0.009 ***		0.027
	(0.02)	(0.30)		(0.02)	(0.00)		(0.02)
I			0.007 **	0.006 *
			(0.00)	(0.00)
T						−2.159 ***	−2.006 ***
						(0.47)	(0.48)
Time effect	YES	YES	YES	YES	YES	YES	YES
Control variables	YES	YES	YES	YES	YES	YES	YES
Constant	−0.144 ***	2.109 **	−0.130 **	−0.156 ***	0.370 ***	0.672 ***	0.597 ***
	(0.06)	(0.88)	(0.05)	(0.06)	(0.01)	(0.18)	(0.18)
R2	0.398	0.972	0.398	0.403	0.941	0.418	0.420
Sobel-Z	-	1.690			3.309
Bootstrap confidence interval	-	[0.0016, 0.0152]			[0.0116, 0.0321]

Note: The standard errors are in parentheses, with *, **, and *** representing significant results at the 0.10, 0.05, and 0.01 levels, respectively.

reports the four-stage estimation results of the mediating effect. Among them, column (1) is the same as column (3) in Table 3, representing the total effect.

Regarding the transmission path of “Spatial agglomeration of construction land → Innovation → Economic resilience”, Column (2) shows that the spatial agglomeration of construction land resources has a significant positive effect on innovation at the 1% significance level, indicating that the spatial agglomeration of construction land resources drives innovation through the learning mechanism. Column (3) shows the impact of innovation on economic resilience, which is significantly positive at the 5% significance level, suggesting that innovation is an important driving force for economic resilience. This is consistent with the conclusions of Pilerdenti, A. et al. (2020) [31] and Bristow, G. et al. (2018) [32]. Column (4) presents the estimation results when both the spatial agglomeration of construction land resources and innovation are included in the determinant model of economic resilience. Innovation is significantly positive at the 10% significance level, while the significance level of the spatial agglomeration of construction land resources decreases compared with the total effect, and the impact intensity decreases by 13.04%. This preliminarily suggests that innovation serves as an important transmission channel through which the spatial agglomeration of construction land resources impacts economic resilience. The Sobel test shows that its Z-statistic is 1.690, significant at the 10% level; the Bootstrap (1000 times) sampling test shows that the confidence interval of the mediating effect with a 95% confidence level does not contain 0, indicating that the test result of the mediating effect is valid. Therefore, Hypothesis 2 holds.

From the perspective of the transmission path of “construction land spatial agglomeration → technical efficiency → economic resilience”. Among them, Equation (5) shows that construction land spatial agglomeration is negatively correlated with technical efficiency at the 1% significance level, but inconsistent with Hypothesis 3. Further analysis reveals that the reason is the significant U-shaped relationship between the spatial agglomeration of construction land and technical efficiency, with the inflection point value being 0.526. In line with China’s actual development, only Yunnan and Guizhou provinces have exceeded the inflection point. The segmented regression at the inflection point shows that, on the left side of the inflection point, the spatial agglomeration of construction land is negatively correlated with technical efficiency at the 1% level. The reason behind this is that the overall spatial agglomeration degree of construction land in China’s eastern and central regions is relatively low, and no obvious provincial spatial agglomeration economic effect has been formed. This results in a significant gap between the actual provincial GDP output and the ideal GDP output. As Professor Lu Ming pointed out in his book Great Country, Great City, China’s cities are not too big, but rather too small, and distorted land policies have led to a decline in economic efficiency. Therefore, it is necessary to drive development with large cities and leverage the advantages of China’s large economy [39]. Equation (6) shows the impact of technical efficiency alone on economic resilience, which is significantly negative at the 1% significance level. This indicates that the current technical efficiency has not led to an increase in economic resilience, which is inconsistent with the expectation of Hypothesis 3. As mentioned above, further analysis in this paper reveals that there is a U-shaped relationship between technical efficiency and economic resilience, but the inflection point value exceeds the range of values considered in this study. This suggests that, at this stage, the overall technical efficiency of Chinese provinces is relatively low, which in turn is not conducive to enhancing economic resilience. In fact, studies have found that, over time, the contribution of total factor productivity to economic growth has been continuously declining [40], and the fluctuations in total factor productivity growth are highly correlated with the fluctuations in GDP. The deterioration of technical efficiency has an inhibitory effect on the growth of total factor productivity [41], which can further lead to the inference that the current technical efficiency is not conducive to economic resilience. The reason is that the overall technical efficiency is relatively low. Column (7) reports the simultaneous impact of construction land spatial agglomeration and technical efficiency on economic resilience. Among them, technical efficiency is significantly negative at the 1% significance level, while the coefficient of construction land spatial agglomeration decreases by 41.30% and is no longer significant. The sign of the product of coefficient b₁ and coefficient c₂ is consistent with the sign of coefficient α₁. It can be preliminarily judged that technical efficiency is another important transmission channel through which the spatial agglomeration of construction land resources impacts economic resilience. The Sobel test shows that its Z-statistic is 3.309, which is significant at least at the 1% level; the Bootstrap (1000 times) sampling test shows that the confidence interval of the mediating effect with a confidence level of 95% does not contain 0, indicating that the test result of the mediating effect is valid. In summary, the spatial agglomeration of construction land can affect economic resilience through technical efficiency.

4.5. Expanded Research

4.5.1. The Regulatory Effect of the Scale of Provincial Construction Land

Generally speaking, provincial primate city is in the position of a growth pole. The larger its scale, the stronger its radiation and driving effect on surrounding cities. The spatial agglomeration of construction land is a relative indicator. For the same degree of spatial agglomeration of construction land, different total scales of provincial construction land will lead to different natures and functions. When the scale of provincial construction land is larger, the scale of construction land resources invested in the primate city will be larger, and the spatial agglomeration of construction land will show high-quality agglomeration, and the primate city will have a stronger ability as a growth pole. When the scale of provincial construction land is smaller, the actual scale of construction land resources invested in the primate city will be relatively smaller, the construction land resources will show low-level spatial agglomeration, and the primate city’s ability as a growth pole will be weaker.

Table 6. Test of moderating effect.

Variable	(1) R_A	(2) R_A	(3) R_A	(4) R_A
S	0.052 ***	0.068 ***	0.046 **	0.061 ***
	(0.02)	(0.02)	(0.02)	(0.02)
scale	0.019 ***	0.019 ***	0.015 *	0.014 *
	(0.01)	(0.01)	(0.01)	(0.01)
S×scale		0.087 ***		0.072 ***
		(0.02)		(0.02)
Time effect	YES	YES	YES	YES
Control variable	NO	NO	YES	YES
Constant	−0.126 ***	−0.123 ***	−0.144 ***	−0.141 **
	(0.04)	(0.04)	(0.06)	(0.06)
R2	0.302	0.322	0.398	0.410

Note: The standard errors are in parentheses, with *, **, and *** representing significant results at the 0.10, 0.05, and 0.01 levels, respectively.

reports the moderating effect of the scale of provincial construction land on the relationship between its spatial agglomeration and economic resilience. Columns (1) and (3) show that, regardless of whether control variables are considered or not, the spatial agglomeration of construction land and the scale of provincial construction land have a significant positive effect on economic resilience. It is worth noting that column (1) shows that when the control variable of provincial construction land scale is added, the strength and significance of the effect of construction land spatial agglomeration on economic resilience are significantly improved compared with column (1) of the benchmark regression in Table 2. This indicates that the provincial construction land scale will, to a certain extent, strengthen the positive effect of construction land spatial agglomeration on economic resilience. Or, without controlling for the provincial construction land scale, the effect of construction land spatial agglomeration on economic resilience will be masked. Based on this, the interaction term of the two is introduced into the model to test this inference. The estimation results reported in columns (2) and (4) show that the interaction term (S×scale) of the two is significant at the 1% level, and the spatial agglomeration of construction land is still significant at the 1% level. Moreover, its effect strength is increased by 30.77% and 32.61% respectively compared with columns (1) and (3), that is, the scale of construction land will strengthen the positive impact of construction land spatial agglomeration on economic resilience.

4.5.2. Non-Linear Effect

The spatial agglomeration of construction land is a dynamic process. Within a certain period of time, it can be successively divided into three stages: under-agglomeration, optimal agglomeration, and over-agglomeration. When the agglomeration is weak or excessive, it may have certain adverse effects on the overall economic resilience of the province. The above mentioned research findings show that the spatial agglomeration of construction land resources only has a promoting effect on economic resilience at the 5% significance level, which is probably the result of the combination of the positive and negative effects of construction land spatial agglomeration. On the basis of the benchmark regression, this paper incorporates the quadratic and cubic terms of the spatial agglomeration of construction land into the model.

The results reported in columns (1) and (2) of

Table 7. Test of nonlinear effects.

Variable	(1) Total Sample	(2) Total Sample	(3) Total Sample	(4) Total Sample	(5) Turning Point Left	(6) Turning Point Right	(7) 0.1 Quantile	(8) 0.25 Quantile	(9) 0.50 Quantile	(10) 0.75 Quantile	(11) 0.90 Quantile
S	−0.253	−0.297 *	0.250 ***	0.288 ***	0.104 ***	0.160	0.016	0.035	0.054 **	0.047 **	0.058 ***
	(0.16)	(0.17)	(0.06)	(0.06)	(0.02)	(0.13)	(0.02)	(0.03)	(0.02)	(0.02)	(0.02)
S2	1.097 ***	1.281 ***	−0.269 ***	−0.305 ***
	(0.42)	(0.43)	(0.08)	(0.08)
S2	−1.117 *** (0.33)	−1.290 *** (0. 34)
Time effect	YES	YES	YES	YES	YES	YES	YES	YES	YES	YES	YES
Control variables	NO	YES	NO	YES	YES	YES	YES	YES	YES	YES	YES
Constant	0.017	−0.093	−0.037 ***	−0.191 ***	−0.115 **	0.073	−0.209 ***	−0.182 **	−0.080	−0.073	−0.082 *
	(0.02)	(0.06)	(0.01)	(0.06)	(0.06)	(0.29)	(0.05)	(0.07)	(0.06)	(0.05)	(0.05)
N	529	529	529	529	456	73	529	529	529	529	529
R2	0.323	0.435	0.308	0.418	0.463	0.716	0.439	0.351	0.344	0.412	0.489

Note: The standard errors are in parentheses, with *, **, and *** representing significant results at the 0.10, 0.05, and 0.01 levels, respectively.

show that, without controlling for other variables, the linear term of the spatial agglomeration of construction land is insignificant. However, controlling for other control variables, the spatial agglomeration of construction land exhibits an inverse N-shaped (И) relationship with economic resilience. Specifically, the coefficient of the linear term is significant only at the 10% level, while both the quadratic and cubic terms are significant at the 1% level. Moreover, further examination of the inflection points reveals that the first inflection point does not exist within the sample range.

Based on this, this paper only incorporates the quadratic term of the spatial agglomeration of construction land into the model on the basis of the benchmark regression, so as to identify the stage heterogeneity of the effect of the spatial agglomeration of construction land on economic resilience. Columns (3) and (4) show that, regardless of whether control variables are added or not, there is a significant inverted U-shaped relationship between the spatial agglomeration of construction land and economic resilience at the 1% level. This indicates that, in the benchmark regression part, the spatial agglomeration of construction land only has a promoting effect on economic resilience at the 5% level, which is very likely the result of the superposition of positive and negative effects. In order to more effectively verify the existence of the inverted U-shaped relationship between the two, the sample is divided into different sample groups, with values less than and greater than the inflection point, and on this basis, the impact of construction land spatial agglomeration on economic resilience is tested. The sub-sample test shows that on the left side of the inflection point, the spatial agglomeration of construction land is significantly positive at the 1% level; on the right side of the inflection point, its effect is not significant. From the perspective of China’s development practice, except for Yunnan and Guizhou provinces, the spatial agglomeration level of construction land in other provincial regions is on the left side of the inflection point value during the sample period. In summary, there is a certain inverted U-shaped driving process between the spatial agglomeration of provincial construction land resources and economic resilience, but its positive external effect dominates.

In order to further reveal the non-linear relationship between the two, this paper uses quantile regression to examine the impact of construction land spatial agglomeration on economic resilience at different quantiles, such as 0.1, 0.25, 0.50, 0.75, and 0.90. The results reported in columns (7)–(11) of Table 7 show that the impact coefficient and significance level increase relatively as the quantile increases, which indicates that there is a non-linear relationship between construction land spatial agglomeration and economic resilience.

4.5.3. Regional Heterogeneity

Based on the evolution law of the spatial organizational structure and the differences in regional development stages, the spatial agglomeration of construction land resources in Chinese provincial regions shows a characteristic of gradually decreasing from west to east. Therefore, this paper divides the total sample into three subsamples: the eastern region (Hebei, Liaoning, Shandong, Zhejiang, Jiangsu, Fujian, and Guangdong), the central region (Shanxi, Henan, Hubei, Anhui, Hunan, Jiangxi, Heilongjiang, and Jilin), and the western region (Shaanxi, Sichuan, Gansu, Ningxia, Yunnan, Guizhou, Guangxi, and Inner Mongolia). On this basis, this paper examines the regional heterogeneity of the impact of the spatial agglomeration of construction land resources on economic resilience.

Table 8. Test of regional heterogeneity effects.

Variable	(1) Eastern Region	(2) Eastern Region	(3) Central Region	(4) Central Region	(5) Western Region	(6) Western Region
S	0.022	−0.769 **	0.041	−0.067	0.069 **	0.533 ***
	(0.06)	(0.37)	(0.05)	(0.15)	(0.03)	(0.11)
S2		1.488 **		0.153		−0.509 ***
		(0.69)		(0.20)		(0.12)
Time effect	YES	YES	YES	YES	YES	YES
Control variable	YES	YES	YES	YES	YES	YES
Constant	−0.289 *	−0.199	−0.055	−0.029	0.011	−0.120
	(0.15)	(0.15)	(0.11)	(0.12)	(0.12)	(0.12)
N	161	161	184	184	184	184
R2	0.652	0.666	0.485	0.487	0.582	0.629

Note: The standard errors are in parentheses, with *, **, and *** representing significant results at the 0.10, 0.05, and 0.01 levels, respectively. The subinflection point of the eastern and western regions has returned because the layout has not been reported.

reports the estimation results.

Regarding the eastern region, column (1) shows that the coefficient of the spatial agglomeration of construction land resources is positive but not significant. The reasons are twofold. On the one hand, the land supply policy is biased towards the central and western regions, which has led to relatively scarce construction land quotas in the eastern provinces. On the other hand, in line with the strategy of balanced regional development, policies have restricted the supply of construction land in large cities. This has caused the distribution of construction land resources in the leading cities of the eastern provinces to be less prominent overall, with a less noticeable growth pole effect. Considering the possible non-linear relationship, column (2) shows that there is a U-shaped relationship between the spatial agglomeration of construction land and economic resilience, which is different from the basic conclusion of the whole country, and it is significant at the 5% level. The inflection point value is 0.258, which indicates that, only when the spatial agglomeration of construction land resources exceeds the inflection point can the positive effects of agglomeration be realized, thereby driving innovative development, promoting industrial upgrading, and consequently contributing to the enhancement of economic resilience. During the study period, only Zhejiang and Fujian are on the right side of the inflection point, which shows that the negative effect of the spatial agglomeration of construction land resources on economic resilience in the eastern provincial regions dominates. The regression by inflection point also confirms this conclusion. As mentioned above, most provinces in the eastern region are facing a shortage of construction land quotas and are in the stage of decentralized spatial agglomeration development; therefore, the insufficient spatial agglomeration of construction land resources in the eastern region makes the negative effect of spatial agglomeration dominant. Under the superposition of positive and negative effects, the overall effect of the spatial agglomeration of construction land resources on empowering economic resilience is not significant.

Regarding the central region, column (3) shows that the effect of spatial agglomeration of construction land is not significant. However, in terms of horizontal comparison, compared with the eastern provincial regions, the spatial agglomeration of construction land in the central provincial regions has a relatively obvious effect on enhancing economic resilience. Further considering the non-linear relationship between the two, column (4) shows that the non-linear relationship does not hold. In summary, the spatial agglomeration of construction land in the central region has not reached the level of empowering economic resilience.

Regarding the western region, column (5) shows that the spatial agglomeration of construction land resources is significantly positive at the 5% level, indicating that the spatial agglomeration of construction land in the western provincial regions has a good effect on empowering economic resilience, which is consistent with the conclusion of the whole sample. The non-linear test reported in column (6) shows that the spatial agglomeration of construction land resources has an inverted U-shaped driving process on economic resilience, and it is significant at the 1% level. The inflection point is 0.524. This indicates that the western region, when provincial resources and elements are overly concentrated in the provincial capital, will generate agglomeration congestion effects, weaken the functional division of labor among cities, and be detrimental to the integration process within the province, thereby suppressing economic resilience [42]. The regression by inflection point shows that on the left side of the inflection point, the spatial agglomeration of construction land has a significant effect on economic resilience at the 1% level, while on the right side of the inflection point, its effect is not significant. In terms of development practice, only Yunnan and Guizhou are on the right side of the inflection point. In summary, the promoting effect of the spatial agglomeration of construction land resources on economic resilience in the western region dominates.

5. Conclusions and Discussion

5.1. Conclusions

How to enhance economic resilience has been a significant issue for Chinese society following the global financial crisis of 2008. Construction land resources have played an important role in China’s economic growth. At the same time, China’s land resource allocation is unique. Based on the methodological approach of structural functionalism and the “Structure-Conduct-Performance” (SCP) analytical framework, this study employs fixed-effects models, mediating effect models, moderating effect models, quantile regression, and subsample regression. Utilizing Chinese provincial panel data since 2000, it examines the effects of the spatial agglomeration of construction land resources on empowering economic resilience and investigates its underlying mechanisms during the spatial allocation process of construction land. The research findings are as follows: First, the spatial agglomeration of construction land has a positive empowering effect on economic resilience. Second, innovation and technical efficiency are important transmission paths for the spatial agglomeration of construction land to empower economic resilience. In comparison, the mediating effect of technical efficiency is stronger. Third, the empowering effect of the spatial agglomeration of construction land on economic resilience shows an inverted U-shaped process, with the promoting effect being dominant. The empowering effect strengthens as the quantile increases and becomes stronger from east to west, consistent with the basic pattern observed nationwide. Fourth, the scale of provincial-level construction land will strengthen the empowering effect of the spatial agglomeration of construction land on economic resilience.

5.2. Recommendation

Based on the results of the empirical research, the following suggestions are put forward.

First, implement and deepen the major function oriented zone strategy. The master functional zone planning is the top-level plan for the overall layout of national land space development and protection at the national and provincial scales for the national land space planning. Building principal functional zones is a major strategy for China’s economic development under the premise of ecological environment protection. At the provincial level, on the one hand, it is necessary to standardize the order of spatial development, strengthen provincial-level national land space control, scientifically and comprehensively delimit the “three zones and three lines”, adhere to the principle of both strict protection and rational use of land resources, and implement the provincial-level national land main functional area strategy. Efforts should be made to combine the “three zones” with the positioning of provincial-level urban development, strengthen the key development and optimized development of urbanized areas, adhere to the direction of centralized spatial allocation of construction land resources in such areas, strengthen the restricted development and prohibited development of agricultural functional areas and ecological functional areas, and reduce the allocation of construction land quotas for them, so as to strengthen the urban functional positioning and urban functional division of labor and improve the overall economic resilience of the province through urban functional division of labor. On the other hand, it is important to deepen the strategic thinking of principal functional zones and construct a “principal functional zone strategy” in the field of construction land resources at the provincial-level urban scale. Provincial government should adhere to the principle of spatial justice. Different cities have absolute but differentiated spatial development rights. Based on the evaluation of resource- environment carrying capacity and the suitability of national land space development, the “principal functional zone” in the field of construction land resources at the provincial-level urban level and the spatial planning system of construction land resources should be formulated.

Second, improve the spatial allocation mechanism of provincial-level construction land resources. Give full play to the leading role of provincial-level governments in the “primary distribution” of construction land. Provincial-level governments should fully respect the basic economic law that mobile production factors such as labor and capital agglomerate in (super) large cities. In the process of spatial allocation of construction land resources, attention should be paid to the spatial coupling and synergy between the agglomeration of factors such as labor and capital and the agglomeration of construction land resources. Construction land quotas should be tilted towards central cities to promote the supply–demand matching of land resources in space, prevent the overall economic efficiency loss caused by the spatial mismatch of human and land resources, guide the spatial functional division of labor through the spatial agglomeration of construction land resources, and give play to the economic resilience effect of spatial functional division of labor. It is important to give full play to the “secondary distribution” role of the market mechanism in the spatial allocation of construction land resources. A sound cross-city trading mechanism and cooperation mechanism should be established for provincial-level urban construction land quotas to drive the “flow” of construction land resources in provincial-level cities to central cities with location advantages and improve the comprehensive carrying capacity of central cities. On the one hand, they should give play to their status as geographical growth poles in the provincial-level economy. On the other hand, central cities should drive to become the sources of provincial-level innovation and development and promote the provincial-level industrialization process so as to enhance the provincial-level economic resilience.

Third, the centralized allocation of construction land quotas to advantageous regions should be strengthened. Research has found that the scale of construction land in a province will enhance the role of spatial agglomeration of construction land in promoting economic resilience. However, sub-sample analysis shows, that in the eastern region, the role of spatial agglomeration of construction land in promoting economic resilience is not significant. This indicates that China should focus on tilting the construction land quotas towards the eastern coastal provinces to provide more construction land quotas for regions with economic growth advantages. Empirical analysis shows that, at the provincial level, different regions have different effects of provincial construction land resources on economic resilience and various spatial agglomeration patterns of construction land. Most of the eastern coastal provinces adopt a decentralized agglomeration pattern, with relatively weak spatial agglomeration of construction land resources, while most of the inland areas adopt a centralized agglomeration pattern, with obvious spatial agglomeration. For the coastal provinces themselves, it is necessary to change the policies that restrict the supply of construction land in large cities and strengthen the centralized allocation of construction land resources to those large cities with advantages. For western provinces, it is necessary to avoid the dual urban diseases of low-efficiency utilization of construction land resources in central cities and shortage of land resources in non-central cities caused by the excessive agglomeration of construction land resources in central cities.

5.3. Discussion

Of course, there are also some limitations and prospects in this study, which are specifically as follows.

First, other measurement methods for the spatial agglomeration of provincial construction land resources can be considered. Due to the limitation of space, this paper focuses on the explained variable of economic resilience. In terms of the core explanatory variable of the spatial agglomeration of construction land resources in China’s provincial regions, this paper only chooses the share of the largest city to represent it. This indicator is simple, clear, and straightforward, but at the same time, it may not fully reveal its spatial agglomeration level. For example, the spatial Herfindahl Index and Zipf Index can also be used for comprehensive identification. Of course, the spatial agglomeration index obtained in this way becomes abstract, which is not conducive to clarifying the location orientation of the spatial allocation of construction land resources in policy analysis.

Second, the identification methods of provincial economic resilience require further research. The identification method of provincial economic resilience in this paper is based on the elasticity concept of microeconomics. To put it simply, it is the organic expansion of the elasticity idea reflecting the relationship between variables in microeconomics to the regional economic resilience, which reflects the spatial economic relationship. Of course, its basis is the premise of spatial autocorrelation, and this paper assumes that there are economic links between this province and other provinces in the country. However, this assumption may be too strict. In other words, some provinces may only have economic links with some other provinces in the country. The provincial economic resilience obtained in this way may not be accurate.

Third, the analysis of the relevant mechanisms of action can be further enriched. In the analysis of the internal mechanism, this paper only starts from the sharing, matching, and learning mechanisms of agglomeration to extract the innovation and technical efficiency mechanisms of the impact of the spatial agglomeration of provincial construction land on economic resilience. There may be other internal mechanisms that need to be further explored. At the same time, when analyzing the moderating effect, this paper only considers the moderating role of the scale of construction land. In fact, in addition to this, marketization and transportation infrastructure also play an important moderating role.

Fourth, the impact of spatial agglomeration of different functional construction land on provincial economic resilience remains to be discussed. This paper only examines the economic resilience effect of the spatial concentration configuration of construction land from the overall perspective of construction land. In fact, there are different types of land use within the construction land, such as industrial land, residential land, and commercial land. The allocation scale of different types of land use is different in cities with different attributes, thus forming a clear scale distribution characteristic within the entire province. So, what impact will the scale distribution of different types of land use have on economic resilience?