Examining the Impact of National Planning on Rural Residents’ Disposable Income in China—The Case of Functional Zoning

Abstract

The growth of rural residents’ disposable income is essential for narrowing the income gap between urban and rural areas and promoting integrated development. This study explores how China’s National Main Functional Zoning Plan influences rural household income through its regulatory impact on construction land expansion. Using data from county−level administrative units across China, the research identified the construction land regulation index as a key mediating variable linking zoning policy to changes in household income. By shifting the analytical perspective from a traditional urban–rural classification to a framework aligned with the National Main Functional Zoning Plan, the study reveals how spatial planning tools, particularly differentiated land quota allocations, influence household income. The empirical results confirm a structured causal chain in which zoning policy affects land development intensity, which in turn drives rural income growth. This relationship varies across different functional zones. In key development zones, strict land control limits income potential by constraining land supply. In main agricultural production zones, moderate regulatory control enhances land use efficiency and contributes to higher income levels. In key ecological function zones, ecological constraints require diverse approaches to value realization. The investigation contributes both theoretical and practical insights by elucidating the microeconomic effects of national spatial planning policies and offering actionable guidance for optimizing land use regulation to support income growth tailored to regional functions.

1. Introduction

With the continuous refinement of China’s national planning system, spatial governance has already shifted from traditional land use regulation to functional zoning [1]. Functional zoning now plays a pivotal role in national governance, reshaping the allocation of land resources and the structure of development opportunities [2]. While no direct global equivalent exists, Western frameworks—such as Germany’s polycentric equilibrium zoning [3], England’s regional revival potential taxonomy [4], and the United States’ federally guided state–metropolitan governance [5]—share conceptual similarities with China’s National Main Functional Zoning Plan. These systems rely on decentralized autonomy and inter−jurisdictional negotiation, contrasting sharply with China’s centralized, hierarchical approach [6]. Under China’s centralized framework, functional zoning serves as a key policy instrument for implementing national planning in rural areas. As such, it influences not only land use patterns and investment flows, but also the level of household disposable income among rural residents.

As a concrete implementation tool of national planning in rural areas, functional zoning influences not only land use patterns and investment flows, but also the level of disposable income among rural households [7]. Although research on national planning and spatial policies related to rural development has increased substantially, including studies that demonstrate positive effects on rural income through regional comparative advantages [8,9,10], infrastructure investment [11], and urban and rural integration [12], most existing studies have primarily focused on regional economic growth or comparisons between urban and rural areas [13,14,15,16,17,18]. There is still a considerable need for empirical investigation into the causal pathways through which functional zoning policies affect household−level rural income, particularly with regard to identifying specific transmission mechanisms.

Several studies have explored the indirect welfare impacts of spatial policies and zoning instruments, such as restricted development zones and ecological redlines [19,20,21,22,23]. These policies influence household living conditions and income potential by reshaping spatial configurations and reallocating development opportunities [18]. In addition, an increasing number of studies have explored specific transmission channels. These include land conversion [13,24,25,26,27], public service provision [28], ecological constraints [29,30], labor mobility [31,32], and institutional environments [33,34,35]. However, these studies have generally concentrated on macro−level structural transformation or ecological externalities, and few have addressed land use control intensity as an institutional constraint embedded within the process of income distribution. This gap underscores the need for further empirical investigation into how functional zoning policies operate through spatial regulatory mechanisms to influence income outcomes at the household level.

In practice, land use control indicators serve as core instruments within the functional zoning system. These indicators determine the spatial boundaries, intensity, and accessibility of rural land use, and they influence agricultural production, the layout of non−agricultural industries, and the direction of investment. As such, land use control represents one of the most operational and restrictive mechanisms linking zoning classifications to income variation. Compared to earlier research that emphasizes structural dualism, infrastructure expansion, or ecological trade−offs, land use control is adopted as a mediating variable in order to better reveal how functional zoning policies affect rural household income through spatial institutional constraints. This approach offers improved theoretical relevance and more robust policy interpretation.

In this study, we aim to explore the specific relationship between functional zoning and the disposable income of rural residents, with particular emphasis on how land use control indicators embedded in functional zoning master plans affect income levels. The research focuses on two key questions: first, whether land use control strength serves as a key mediator in the potential impact of functional zoning on rural residents’ disposable income; and second, how different types of land use control specifically affect income. Methodologically, this research employs the stepwise regression approach proposed by Baron and Kenny and utilizes the bootstrap method to test the significance of the mediation effect. For the first time, land use control strength is introduced as a mediator in this study, constructing a chain−like transmission mechanism from functional zoning to land control and ultimately to rural residents’ disposable income. This analysis addresses the critical research gap regarding the impact of national planning on rural income through institutional land constraints. It further deepens the understanding of regional differences in the effects of land total control on income in spatial planning, contributing to the development of an income adjustment framework within regional governance under the rural revitalization strategy.

The analysis is structured to first present the theoretical framework and research hypotheses, followed by a detailed methodology section covering variable definition, robustness checks, mediation models, and the Geographically Weighted Random Forest (GWRF) approach. The paper then proceeds with the results, discussion, and conclusions, offering insights into mediation model findings, robustness checks, GWRF outcomes, and their implications for rural income growth.

2. Theoretical Analysis Framework

2.1. The Impact of Main Functional Zones on the Disposable Income of Rural Residents

The classification of main functional zones (See

Table 1. Conceptual framework of China’s functional zones.

Main Functional Zone Type	Major Function	Minor Function	Description	Policy Objective
Key Development Zones	Massive urbanization and industrialization	Food production, ecological construction, etc.	Regions with ample carrying capacity and rapid potential for industrialization and urbanization, designated as “potential areas”.	Emphasis on economic scale growth
Optimized Development Zones	Massive urbanization and industrialization	Food production, ecological construction, etc.	High−density urban “cores” whose environmental carrying capacity is near its limit.	Emphasis on improving development quality
Main Agricultural Production Zones	Food production	Ecological construction, moderate industrialization and urbanization	National “granaries” whose primary task is to secure grain and major agricultural products.	Emphasis on ensuring food security
Key Ecological Function Zones	Ecological safety	Ecological economy	National “green lungs” that provide ecological security barriers and ecosystem services.	Emphasis on ecological protection priority

) significantly shapes the allocation of resources in rural economic activities by applying differentiated strengths of land use control, which may in turn influence the disposable income of rural residents.

In optimized development zones, identified as high−density urban “cores” with nearly exhausted environmental carrying capacity, tasked with massive urbanization and industrialization alongside minor functions like food production and ecological construction, the policy objective emphasizes improving development quality [2,36]. The focus on industrial upgrading and intensive land use is accompanied by strict land control, which restricts non−farm development and may lead to suppressed income growth for rural residents due to limited diversification opportunities.

In key development zones, designated as regions with ample environmental carrying capacity and rapid potential for massive urbanization and industrialization, with minor functions in food production and ecological construction, the policy objective prioritizes economic scale growth [7]. More flexible land policies allow for moderate development and industrial promotion, which may contribute to increased transfer income among rural residents through expanded economic opportunities.

Main agricultural production zones serving as national “granaries”, with a primary task of securing grain and major agricultural products, while supporting ecological construction and moderate industrialization and urbanization, aim to ensure food security as their policy objective [37]. These zones apply moderately strict land control, primarily oriented toward agricultural infrastructure, which limits land use diversification but may support agricultural income enhancement through investments in farming productivity.

In key ecological function zones, acting as national “green lungs” that provide ecological security barriers and ecosystem services, with a minor focus on ecological economy and a policy objective of prioritizing ecological protection, the strictest land control is enforced to prohibit large−scale development [1,38]. This prioritization of ecological conservation constrains rural industrialization, which may result in contracted income opportunities for rural residents due to restricted economic activities (See Figure 1).

Overall, land control policies embedded in different functional zoning types shape land use patterns, which in turn may influence the rural economic development model by either facilitating or limiting the potential for rural residents’ income growth to varying degrees.

2.2. Pathways from Land Use Control to Income Outcomes

Land use control indicators influence rural residents’ income through multiple interconnected pathways that reflect the regulatory effects of functional zoning. First, land use control intensity imposes direct restrictions on land development, particularly by limiting the availability of land for non−agricultural industries in rural areas, such as township enterprises. This reduction in development capacity leads to fewer local employment opportunities, thereby slowing the growth of rural residents’ income. Second, in regions with limited land development and fewer labor−intensive industries, the rural labor force is more likely to migrate elsewhere for work. This increase in outbound migration makes rural households more dependent on external sources of income, directly affecting their disposable income. Third, stringent land control policies may deter external investment. By reducing the feasibility of large−scale investment projects, such as factories and industrial enterprises, the intensity of land control lowers the overall level of investment, thereby constraining local economic development and further influencing household income levels. Fourth, restrictive land use policies can act as a barrier to the entry of leading enterprises. Large firms, particularly those involved in agricultural product processing, often generate local employment and stimulate rural economic activity. When these enterprises are unable to establish themselves due to land control constraints, the absence of their economic spillover effects can further limit increases in rural household income (See Figure 2).

In summary, land use control affects rural residents’ income through the joint operation of multiple pathways. The strength and direction of the impact vary depending on the type of functional zone to which a region is assigned under the national spatial planning system.

2.3. Research Hypothesis

Building upon the analytical framework established in Section 2.1 and Section 2.2, which delineates how functional zoning shapes land use control and its subsequent pathways, influencing rural residents’ income, we formalize the following research hypotheses:

Hypothesis 1 proposes that land use control strength serves as a mediating factor between functional zoning and rural household income. The degree of control over land development reflects the extent to which local economic activities are permitted or restricted. This control strength transforms the policy orientation of functional zones into tangible changes in household income, thereby acting as a key linkage between spatial policy categorization and income variation at the household level.

Hypothesis 2 suggests that different types of functional zones have significant and distinct effects on the income levels of rural residents. These differences mainly result from variations in the policy orientations of each zone, including aspects such as resource allocation, industrial planning, ecological protection goals, and development priorities. In addition, the impact of functional zoning is influenced by geographical and regional conditions, leading to differentiated outcomes across various areas.

3. Methods

3.1. Variable Definition

3.1.1. Dependent Variable

The dependent variable in this study is the disposable income of rural residents, defined as the income remaining for consumption and savings after mandatory deductions such as taxes and social security contributions. This indicator is widely adopted and readily measurable, offering a reliable reflection of the economic status and welfare of rural populations. It serves as a key metric for evaluating rural development, household well−being, and the effectiveness of policy measures [39,40].

The data on rural disposable income for the year 2020 were collected at the county level, based on the China County Statistical Yearbook, a highly authoritative source known for its standardized and comprehensive economic statistics. This ensures data consistency and comparability across different regions.

As illustrated in Figure 3, which shows the spatial distribution of per capita rural disposable income in 2020, income levels are generally higher in the eastern and southeastern regions, while the northern and western regions exhibit significantly lower levels. This pattern reveals the substantial regional disparities in rural economic development across China.

3.1.2. Core Independent Variable

The core independent variable in the experiment is control strength, a proxy for the regulatory intensity linked to different types and administrative levels within main functional zones. The Main Functional Zoning Plan is the foundational framework that guides China’s territorial development and ecological protection. It divides the national territory into four functional categories: optimized development zones, prioritized development zones, main agricultural zones, and key ecological function zones. This framework promotes the orderly allocation of resources across regions and optimizes the national spatial pattern [41]. Each zone type carries a distinct development focus: optimized development zones encourage industrial upgrading and efficient resource use in densely populated areas; prioritized development zones support infrastructure expansion and the clustering of population and industry; main agricultural zones safeguard food security through farmland preservation and agricultural modernization; and key ecological zones emphasize environmental conservation by strictly limiting development intensity.

To measure differences in land use regulation across these zone types and administrative hierarchies, namely national and provincial levels, we employ control strength as a numeric index. Following the classification scheme of Guo et al. [42], each zone is assigned a value from 1 to 8, as shown in

Table 2. Control strength by main functional area type.

Main Functional Zone Type	Administrative Level	Control Strength
Key Development Zones	National	1
Key Development Zones	Provincial	2
Optimized Development Zones	Provincial	3
Optimized Development Zones	National	4
Main Agricultural Production Zones	Provincial	5
Main Agricultural Production Zones	National	6
Key Ecological Function Zones	Provincial	7
Key Ecological Function Zones	National	8

. Larger values represent stronger regulatory constraints and a more limited capacity for construction land expansion.

3.1.3. Mediating Variable

The mediating variable in this study is the construction land regulation (CLR) index, defined as the ratio of the increase in total construction land area (CLA) from 2005 to 2020 to the base−year area in 2005. This index reflects the relative change in land development intensity over time at the regional level. The formula for this index is

$$CLR Index={\displaystyle \frac{{CLA}_{2020}{-CLA}_{2005}}{{ CLA}_{2005}}}$$

(1)

This formula calculates the increase in CLA (the difference between 2020 and 2005) relative to the base−year area in 2005, providing a quantitative measure of development intensity. The theoretical rationale for treating this indicator as a mediating variable is robust.

The theoretical rationale for treating this indicator as a mediating variable is robust. First, as an essential policy tool of China’s national spatial governance system, the classification of major functional zones imposes systematic differences in land use regulation and the delineation of development boundaries. For instance, optimized development zones tend to have greater land use flexibility, whereas ecological function zones are subject to stricter controls. These institutional distinctions influence the pace and scale of construction land expansion across regions.

Second, the growth of construction land serves as a critical indicator of land resource allocation and development dynamics. Increased development intensity is often associated with infrastructure improvements, the introduction of industrial and service−sector investments, and the expansion of local labor markets, which together may enhance rural households’ off−farm employment opportunities and income levels. Conversely, overly rigid land use controls may constrain regional economic potential and limit income−generating channels for rural populations.

Thus, the CLR index functions as a meaningful mediator that helps clarify the mechanism through which spatial governance policies influence rural economic outcomes.

3.1.4. Control Variables

We incorporate a set of control variables grounded in classical production−factor theory, encompassing the three key dimensions of labor, land, and capital. These variables collectively reflect underlying regional development conditions and enhance the explanatory power of the empirical model. All control variables are based on data for the year 2020.

For the labor dimension, the year−end total population serves as a proxy for labor supply. A larger population suggests a more abundant potential workforce, which may expand income−generating capacity by supporting industrial and service−sector development.

For the capital dimension, two indicators are employed: total annual fixed−asset investment and the number of newly registered enterprises. Fixed−asset investment reflects the intensity of capital input and overall development potential [39], whereas new business registrations signal market vitality and economic openness. A higher enterprise count typically implies more diversified income channels and a broader non−agricultural employment base [43].

For the land dimension, several factors are included. Farmland area serves as a proxy for agricultural resource endowment, directly supporting land−based income generation [39]. Terrain ruggedness accounts for natural geographic constraints: steep or uneven topography can restrict industrial development, raise infrastructure costs, and hamper accessibility, all of which may limit local economic opportunities [40]. County land area reflects the spatial scale of each unit, influencing development capacity, administrative cost, and land−supply elasticity, thereby indirectly affecting household income [44].

Collectively, these control variables provide a comprehensive picture of regional factor endowments and development environments. Their inclusion helps isolate the specific effects of spatial governance policies and construction−land regulation on rural income outcomes.

3.2. Robustness Checks

To address potential issues of causal inference and endogeneity, we implement robustness checks following the replacement of the dependent variable from rural disposable income to ln_per_capita_gdp and the core independent variable from control strength to zone_type, a categorical variable representing functional zoning types. First, we verify the statistical significance of ln_per_capita_gdp and zone_type across all models to ensure consistent results. Second, we test the mediation effect by assessing the reduction in control strength coefficients after including the mediator ln_land_total_rate, with significance confirmed using 1000 bootstrap resamples. These checks ensure the reliability of the causal pathway from functional zoning to regional economic output via land development intensity.

3.3. Mediation Models

To investigate the impact of control strength associated with various types of regional zoning on rural household income through the CLR index (land total rate), we adopt a mediation analysis framework, wherein the land total rate serves as the mediator linking control strength to rural income. This mediation model is selected for its ability to effectively disentangle the direct and indirect effects of zoning control strength on rural income, enabling us to isolate the mediating role of land use changes while offering a structured approach to examine the causal pathways connecting spatial policies to economic outcomes [45].

To begin, we develop the pathway model to assess the effect of control strength on the CLR index, as detailed in the following equation:

$$ln\_land\_total\_rate\_i={\alpha }_0+{\alpha }_1\ast control\_strength\_i+X\_i\gamma +\epsilon \_i$$

(2)

In this equation, α₀ represents the constant term, which indicates the expected value of the dependent variable when all independent variables are zero. α₁ is the coefficient for control_strength, indicating the effect of the control intensity of regional zoning types on the CLR index. X_iγ represents the control variables, including population size, terrain ruggedness, investment level, farmland area, county area, and the number of new firm registrations, where γ denotes the corresponding coefficients. ε_i is the error term, capturing the influence of unobserved factors.

Next, we construct the no−mediation outcome model to examine the effect of the control intensity of regional zoning types on rural income, with the following equation:

$$ln\_income\_i={\beta }_{0+}{\beta }_1\ast control\_strength\_i+X\_i\delta +\eta \_i$$

(3)

Here, β₀ is the constant term, which indicates the expected value of rural income when all independent variables are zero. β₁ is the coefficient for control_strength, representing the direct effect of the control intensity of regional zoning types on rural income. X_iδ represents the control variables, with δ being the coefficients for these variables. η_i is the error term, accounting for the unobserved factors that influence rural income.

Finally, we construct the mediation outcome model, which includes both the mediator and the independent variable, to examine the effect of regional zoning control intensity on rural income, as specified in the following equation:

$$ln\_income\_i={\theta }_0+{\theta }_1\ast ln\_land\_total\_rate\_i+{\theta }_2\ast control\_strength\_i+X\_i\phi +\mu \_i$$

(4)

In this model, θ₀ is the constant term. θ₁ represents the coefficient for the mediator ln_land total rate, showing the effect of land control on rural income. θ₂ is the coefficient for control_strength, indicating the effect of regional zoning type on rural income after controlling for the mediator. X_iφ represents the control variables, with φ being the coefficients for these variables. μ_i is the error term, accounting for unobserved influences.

All regressions use robust standard errors. To test the mediation effect, we examine whether the coefficient for control strength decreases in magnitude or significance after the mediator is introduced. Additionally, we use 1000 bootstrap resamples to validate the indirect effect. Furthermore, we construct a marginal effects model to quantify the boundary effects associated with different types of main functional zones, aiming to identify zone−specific sensitivities in the relationship between regional zoning control intensity and rural income.

3.4. Geographically Weighted Random Forest (GWRF) Model

Building upon the mediation analysis that identifies a significant pathway from the national zoning policy through CLR to rural household income, we further extend the methodological framework to capture spatial heterogeneity in this causal mechanism. While the mediation model provides insights into average structural effects, it assumes parameter homogeneity across space, potentially obscuring local variations arising from differences in development endowments, institutional enforcement, and land use contexts.

To address this gap, we adopt the Geographically Weighted Random Forest (GWRF) model, selected for its capacity to tackle the spatial heterogeneity prevalent in China’s diverse regional landscapes. This approach integrates spatial weighting with machine learning, enabling it to capture localized effects that traditional models might overlook [46], while also revealing regional variations in the impact of zoning policies on income by allowing parameters to vary spatially, thereby improving the precision of policy implications [47].

Leveraging this framework, we enhance the GWRF approach with Shapley Additive Explanations (SHAP) to improve interpretability and focus on the localized policy impacts across regions. GWRF facilitates the adaptation of model parameters to diverse geographic contexts while effectively capturing nonlinear dynamics and complex variable interactions [48]. We apply the GWRF model to predict rural per capita disposable income as a function of zoning categories, land regulation intensity, and a set of economic, demographic, and locational covariates. We compute spatial SHAP values for each county, enabling us to identify both the direction and strength of each variable’s impact across different locations.

Additionally, to deepen our spatial analysis, we perform a hotspot and coldspot analysis of the SHAP values for the land total rate. By incorporating spatial clustering statistics such as the Getis–Ord Gi index, we identify regions where the influence of this variable on income prediction is significantly concentrated, either as high−value clusters (hotspots) or low−value clusters (coldspots) within different types of main functional zones. This process allows us to pinpoint areas where land development intensity has a notably positive effect on income outcomes.

4. Results

4.1. Mediation Model Findings

Table 3. Regression results of the mediation models.

Variable	Model (1): Ln_land_total_rate	Model (2): ln_income	Model (3): ln_income (with Mediator)
Control strength	−0.0549 **	−0.0305 ***	−0.0277 ***
Ln_population	−0.0279	0.1000 ***	0.1054 ***
Terrain_ruggedness	0.1349 **	−0.0701 ***	−0.0772 ***
Ln_investment	0.0372	0.0181 **	0.0184 **
Ln_farmland	−0.3841 ***	−0.0834 ***	−0.0784 ***
Ln_new_firm_registrations	0.1588 **	0.1322 ***	0.1198 ***
County_area	−0.0005	0.0004 **	0.0004 **
Constant	0.9829	9.5574 ***	9.6235 ***
Ln_land_total_rate			0.0404 ***

Significance codes: *** p < 0.01, ** p < 0.05.

reports the results of the mediation analysis. Model 1 examines the relationship between the independent variable and the mediator (land total rate), revealing that control strength, terrain ruggedness, farmland, and entrepreneurial activity directly affect the growth of land resources. In particular, control strength (−0.0549, p < 0.05) and farmland area (−0.3841, p < 0.01) exhibit significant negative effects, indicating that rational land development and management are essential.

Model 2 further analyzes the relationship between various factors and rural residents’ disposable income. Control strength (−0.0305, p < 0.01) shows a significant negative impact on income, suggesting that strict regulation may suppress local economic activity and income growth by limiting development, resource allocation, and economic freedom. In contrast, population size (0.1000, p < 0.01), investment (0.0181, p < 0.05), and the number of newly registered firms (0.1322, p < 0.01) positively affect income through stimulating economic activity and increasing market demand. Notably, the positive effect of entrepreneurial activity reflects its role as a key driver of income growth. Meanwhile, terrain ruggedness (−0.0701, p < 0.01) and farmland area (−0.0834, p < 0.01) negatively affect income, possibly due to limitations in resource accessibility or land use efficiency.

Model 3 introduces the land expansion rate as a mediator between control strength and income, revealing the transmission mechanism of land resources in this relationship. Specifically, the land total rate (0.0404, p < 0.01) significantly impacts income, indicating that it plays a key mediating role. The bootstrap test indicates that the indirect effect is statistically significant (p = 0.011), providing evidence for a mediation effect whereby the control strength of zone type affects rural income indirectly through its influence on the CLR index. This implies that in regions with stronger control, limited land expansion can constrain income growth.

Control strength exerts an indirect influence on income levels through its impact on the expansion of land resources. However, marginal effect analysis reveals that this transmission mechanism varies across regions—namely, the “width” of the policy effect channel differs depending on regional functional zoning. In other words, even with comparable levels of land expansion, the marginal returns to income are modulated by the specific functional orientation of each zone.

In key development zones, increased control intensity correlates with lower predicted income, suggesting that excessive regulation may hinder economic activity and developmental potential. Conversely, in main agricultural production zones, stronger controls are positively associated with income, likely due to enhanced protection and efficient use of agricultural resources. In key ecological function zones, control intensity negatively affects income by restricting land development and economic activity (see Figure 4). These results highlight that the effectiveness of land use policies depends not only on enabling expansion, but also on the capacity of regions to translate land expansion into economic gains.

4.2. Robustness Checks Results

To ensure the robustness of the mediation analysis, additional tests were conducted by updating the dependent and core explanatory variables, as presented in

Table 4. Regression results of the mediation models with updated dependent variables.

Variable	Model (1): ln_land_total_rate	Model (2): ln_ per_capita_gdp	Model (3): ln_per_capita_gdp (with Mediator)
Control strength	−0.0421 **	−0.0535 ***	−0.0511 ***
Ln_population	0.4492 ***	−0.1288 *	−0.1824 **
Terrain_ruggedness	−0.0622	−0.1060 ***	−0.1079 ***
Ln_investment	0.0440 *	0.0358 *	0.0335 *
Ln_farmland	−0.0716	−0.0882 **	−0.0745 *
Ln_new_firm_registrations	0.1926 ***	0.2464 ***	0.2265 ***
County_area	0.0003	0.0011 ***	0.0011 **
Constant	4.8718 ***	10.9642 ***	10.5371 ***
Ln_land_total_rate			0.0858 **

Significance codes: *** p < 0.01, ** p < 0.05, * p < 0.10.

and

Table 5. Regression results of the mediation models with updated core explanatory variables.

Variable	Model (1): ln_land_total_rate	Model (2): ln_income	Model (3): ln_income (with Mediator)
Zone_type	−0.1292 **	−0.0942 ***	−0.0872 ***
Ln_population	0.0011	0.1148 **	0.1180 **
Terrain_ruggedness	0.1511 **	−0.0580 ***	−0.0665 ***
Ln_investment	0.0387	0.0184 **	0.0185 **
Ln_farmland	−0.4123 ***	−0.1054 ***	−0.0983 ***
Ln_new_firm_registrations	0.1679 **	0.1292 ***	0.1163 ***
County_area	−0.0005	0.0005 **	0.0005 **
Constant	1.1892	9.8540 ***	9.9057 ***
Ln_land_total_rate			0.0410 **

Significance codes: *** p < 0.01, ** p < 0.05.

. Table 4 replaces the dependent variable, the disposable income of rural residents, with GDP per capita to assess whether the mediation effect of the CLR index persists under an alternative economic outcome measure. The results indicate that control strength maintains a significant negative effect on both the mediator and GDP per capita, with coefficients of −0.0421 at the 5% significance level in Model 1, −0.0535 at the 1% significance level in Model 2, and −0.0511 at the 1% significance level in Model 3. The mediator, CLR index, shows a significant positive effect with a coefficient of 0.0858 at the 5% significance level in Model 3. The bootstrap test for the mediation effect yields a p−value of 0.083, indicating marginal significance and suggesting that the mediating role of land use control remains consistent, though less pronounced, when using GDP per capita as the outcome variable.

Table 4 replaces the core explanatory variable, control strength, with a categorical measure of zone type to test the sensitivity of the mediation effect to an alternative specification of functional zoning. The results confirm that the zone type has a significant negative effect on both the mediator and rural income, with coefficients of −0.1292 at the 5% significance level in Model 1, −0.0942 at the 1% significance level in Model 2, and −0.0872 at the 1% significance level in Model 3. The CLR index maintains a significant positive effect with a coefficient of 0.0410 at the 5% significance level in Model 3. The bootstrap test for the mediation effect yields a p−value of 0.008, providing strong evidence of a significant mediation effect. These findings corroborate the primary results, demonstrating that the mediating role of land use control is robust across different specifications of the dependent and explanatory variables, reinforcing the causal pathway from functional zoning to rural income through the CLR index.

4.3. GWRF Model Findings

Figure 5 illustrates the spatial variation in the impacts of control strength and Ln_land total rate on rural per capita disposable income at the national level. Figure 5a presents the SHAP values of control strength, with notable positive contributions concentrated in economically developed southern regions such as the Yangtze River Delta and the Pearl River Delta. This indicates that stricter land use control policies in these regions have had a significant influence on income levels. Figure 5b displays the SHAP values for the Ln_land total rate, revealing that the positive effects of land expansion exhibit clear spatial clustering.

Figure 6 elucidates the spatial heterogeneity in the contribution of the natural logarithm of land total rate to rural disposable income, as shown in hotspot–coldspot distributions across the main functional zones. In main agricultural production zones, as shown in Figure 6c, intense hotspots are concentrated in the Yangtze River Basin provinces of Hunan, Jiangxi, and Anhui. These regions, characterized by fertile alluvial plains and robust institutional support for large−scale farming, demonstrate that land expansion significantly enhances rural income by boosting agricultural productivity and supporting agribusiness, including rice, tea, and aquaculture exports. In contrast, key ecological function zones, depicted in Figure 6d, exhibit persistent coldspots in ecologically sensitive areas, notably the mountainous terrain along the Shanxi−Hebei border, including the Taihang Mountains. Here, stringent conservation policies under China’s Ecological Red Line, coupled with rugged terrain and limited arable land, severely restrict land expansion, thereby constraining its income−enhancing potential. In key development zones and optimized development zones, as illustrated in Figure 6a,b, land expansion displays fragmented hotspot–coldspot patterns, with isolated hotspots in coastal urban clusters and coldspots in peri−urban hinterlands. This variability reflects inconsistent marginal returns to income, driven by competitive land markets, uneven urbanization, and institutional fragmentation in land governance, including conflicts between local and provincial quotas.

Overall, the effects of land expansion on rural income exhibit significant variation across functional zones, shaped by regional differences in market maturity, institutional mandates, and geographical conditions. The analysis highlights that effective policies must convert functional zoning principles into enforceable, spatially grounded regulations. Agricultural zones require measured expansion allowances tied to productivity metrics, while ecological zones demand monetized conservation incentives benchmarked to income alternatives. These tailored interventions transform abstract policy directives into operational parameters, ensuring precision−calibrated outcomes that maximize economic benefits while addressing spatial heterogeneity.

5. Discussion

5.1. Understanding Mediation Effects: Theoretical Insights and Policy Guidance

Identifying the CLR index as a central mediating variable in the relationship between functional zoning policy and rural household income represents a significant theoretical contribution. The immediate restrictions on land use under the National Main Functional Zoning have been shown to widen rural income disparities by limiting development opportunities in ecologically functional zones [49]. In contrast, establishing National Key Ecological Functional Areas within this framework can increase per capita rural income through targeted ecological compensation and support mechanisms [50]. Additionally, the policy demonstrates positive impacts on rural residents’ income growth while reducing ecological poverty, albeit with potential economic tradeoffs [51]. Further studies highlight how such zoning directly enhances local residents’ income, with stronger effects in regions like western and northeastern China [52]. Building on these foundations, which primarily emphasize direct effects, our analysis extends the literature by exploring mediating mechanisms, yielding a more comprehensive grasp of policy dynamics. Moreover, this study advances regional heterogeneity analysis by shifting from a conventional urban–rural dichotomy to a policy−oriented framework aligned with China’s National Main Functional Zoning Plan. This refinement deepens insights into how national spatial planning policies, via tangible instruments like land use quotas, shape household−level economic welfare while bridging broad spatial governance and local income distribution mechanisms.

By empirically validating the transmission pathway from functional zoning control to land development intensity and subsequently to income variation, this study provides the first systematic causal interpretation of land policy impacts on income. In key development zones, strict land regulation reduces land development intensity, indirectly limiting the potential for income growth. In contrast, in main agricultural production zones, moderate regulation optimizes land resource allocation, enhances agricultural productivity, and contributes to increased rural income.

Based on these findings, the study offers the following policy implications: land total rate should serve as a core regulatory variable, with differentiated land development intensity standards tailored to the functional positioning of each zone. In key development zones, moderate relaxation of regulatory constraints is recommended to enhance land allocation efficiency and promote industrial and employment linkages. This can be achieved through several measures, for example, implementing preallocation mechanisms for land quotas, introducing employment−linked approval criteria, and promoting the market−oriented use of rural collective construction land. In main agricultural production zones, while maintaining strict farmland protection, policy efforts should focus on improving the agricultural value chain. This may include, for example, adopting a combination of negative lists and incentive−based measures, and exploring innovative tools such as transferable development rights, to enhance both productivity and farmer income. In key ecological function zones, stronger ecological constraints should be complemented by diversified approaches to ecological value realization. For example, low−impact, site−specific development projects can be piloted; ecological compensation mechanisms can be strengthened; and green skills training and labor mobility programs can be expanded to support a synergistic balance between ecological protection and rural income enhancement.

While these findings provide robust insights into the mediating role of land use control, augmented by the GWRF model’s ability to capture spatial heterogeneity through geographically weighted relationships and SHAP value analysis, certain limitations must be acknowledged to contextualize the results and guide future research. The study relies on cross−sectional data for 2020, with the CLR index derived from changes between 2005 and 2020. This approach captures long−term trends but restricts the ability to model temporal dynamics or lagged policy effects, which could reveal additional nuances in how zoning policies influence income over time. The core independent variable, control strength, is a quantified categorical measure of policy intensity, which, while effective, may not fully account for variations in local policy implementation or unobserved regional factors. Although the mediation analysis, combined with GWRF’s spatial insights, establishes a clear causal pathway, it assumes unidirectional causality, potentially oversimplifying complex feedback loops. Future studies could enhance causal inference by leveraging panel data to capture temporal effects or exploring alternative modeling approaches to further elucidate the interplay between spatial governance and income outcomes in diverse regional contexts.

5.2. Empirical Validation and Policy Implications of the Mediating Role of Land Expansion

Furthermore, recent explorations in local policy implementation across China have provided empirical support for the core mechanism identified in this study, whereby functional zoning control policies affect rural household income indirectly through changes in the CLR index.

In key development zones such as Gan County in Jiangxi Province, flexible regulatory approaches such as the “mining land combined entry into the market” model have alleviated the suppressive effects of strict land controls by releasing quotas for collective construction land [53]. The first transaction under this scheme in 2024 attracted an investment of 95 million RMB, confirming that activating the mediating variable of land total growth rate can significantly promote rural income growth. In main agricultural production zones such as Siling Town in Guangxi, the implementation of the “field chief system” together with sugarcane field consolidation has increased per acre yields by 67 percent [54]. This example demonstrates how productive land regulation can improve land use efficiency while simultaneously supporting farmland protection and agricultural income enhancement.

Further supporting evidence is found in Hubei Province, where a tiered incentive scheme awarded differentiated subsidies of 8 million and 5 million RMB based on cultivated land area [55]. This directly reflects the core finding of this study regarding the mediating role of functional zoning in shaping the relationship between land expansion and income growth.

These practices suggest that future land use policies should focus on improving the transformation efficiency of land total growth rate. In key development zones, this includes enhancing market−based allocation mechanisms [56,57], such as enabling cross−regional quota transfers. In agricultural production zones, efforts should focus on strengthening integrated regulatory and incentive mechanisms [58]. Ultimately, precise zoning based regulation of this key mediating variable is essential to achieving a coordinated balance between land resource protection and rural income enhancement [59].

6. Conclusions

The land total growth rate has been identified as a key mediating variable through which functional zoning policies influence rural household income, offering new insights into the spatial transmission mechanisms of land regulation effects.

(1)

The study demonstrates that functional zoning policies affect rural income not directly, but via changes in land development intensity. In key development zones, strict land control suppresses land expansion and indirectly constrains rural income growth, whereas in main agricultural production zones, moderate regulation improves land use efficiency and supports income enhancement.

(2)

By integrating spatial analysis into the framework of China’s National Main Functional Zoning Plan, this study effectively bridges the gap between macro−level spatial planning and micro−level income distribution. It provides empirical validation for a structured causal pathway wherein functional zoning policies influence land development intensity, which in turn affects rural household income. However, the use of cross−sectional 2020 data limits modeling of temporal dynamics, and future research could employ panel data to better capture long−term policy impacts on income.

(3)

Policy implications derived from the findings emphasize the need for differentiated land development intensity standards based on functional zoning type. Specifically, flexible quota management and employment−linked land allocation in development zones, value chain optimization and development rights trading in agricultural zones, and diversified ecological value realization strategies in ecological zones are key to achieving coordinated rural development.