Research on the Impact of the Reform of “Three Plots of Land” in the Yellow River Basin on Food Security

Abstract

The Yellow River Basin serves as China’s core food security zone and a vital ecological barrier. However, while the “three plots of land” reform has revitalized land resources, it has also exerted complex effects on the allocation of grain production factors. Scientifically assessing the actual impacts of this policy reform on food security and identifying optimization pathways has become a critical issue for safeguarding national food security. Using panel data from 101 county-level administrative units in the Yellow River Basin covering 2010–2023, this study employs a difference-in-differences model and a moderation effect model to systematically evaluate the impact of the “three plots of land” reform policy on food security. By introducing new-type urbanization and agricultural modernization as moderating variables, it further reveals the regional heterogeneity of the policy’s operational mechanisms. The study finds that (1) the “three plots of land” reform policy significantly enhances food security levels, (2) both new-type urbanization and agricultural modernization positively amplify policy effects through moderation mechanisms, and (3) regional heterogeneity tests considering geographical location and climate conditions reveal a spatial gradient pattern of “midstream > downstream > upstream” in policy effects, clarifying the logic of regional heterogeneity. Accordingly, the “three plots of land” reform policy in the Yellow River Basin should be deepened by formulating differentiated policies based on regional heterogeneity. A moderation mechanism should be established where agricultural modernization and new urbanization synergistically support food security, comprehensively enhancing food security safeguarding capabilities.

1. Introduction

The “three plots of land” in rural areas are the “life roots” of hundreds of millions of farmers, which are closely linked to national food security, rural industrial development, and rural harmony and stability. At present and in the coming period, we should continue to deal with the relationship between farmers and land as the main line of rural reform, and focus on grasping the “three plots of land”. The Third Plenary Session of the 20th Central Committee of the Communist Party of China proposed that national security is an important foundation for the stability of Chinese-style modernization [1]. Among them, food security is a matter of national strategic importance, and it is the foundation for safeguarding national security and ensuring economic development. Nine provinces and regions along the Yellow River Basin have accelerated the high-quality development of the Yellow River Basin. The arable land area and total grain output of the Yellow River Basin account for more than 30% of the country, and coal production accounts for about 80% of the country. The role of “ballast stone” in food and energy security continues to be highlighted. Firstly, the construction of a modern industrial system and the cultivation of advanced manufacturing clusters and national strategic emerging industry clusters should be accelerated, while the greening, digitalization, intelligentization, and integration of industries are further promoted [2]. The investment in research and experimental development of the Yellow River Basin accounts for 20.56% of the country’s total investment, and the growth rate exceeds the national average [3]. Secondly, the layout of energy development must be optimized, with the construction of large-scale wind power photovoltaic bases in the upstream “Shago Wasteland” area already being steadily promoted. Currently, the installed capacity of non-fossil energy power generation accounts for more than 48% of the total installed capacity. Thirdly, the development of advantageous agriculture, the continuous optimization of the structure of the agricultural industry, and the construction of advantageous characteristic industrial clusters and strong agricultural industries are all necessary. Their establishment has achieved remarkable results, with 19,800 green, organic, and geographically marked agricultural products in the basin, accounting for about 1/3 of the national total (Chinese People’s Network, 2024). Organic agricultural products constitute the matrix of high-quality development of agriculture. This multiple-approach strategic superposition of “grain–energy–ecology” makes the allocation of land resources a core proposition of watershed development.

The policy effect of the “three plots of land” reform is the focus of academic research [4]. Existing research has confirmed that the “separation of powers” of contracted land promotes the circulation of agricultural land and large-scale operation [5], the paid withdrawal of housing estates can revitalize idle resources [6], and the entry of collective management construction land into the market can solve the bottleneck of rural industrial capital [7]. However, most studies focus on the assessment of a single type of land reform [8], there is a notable lack of systematic analysis of the reform of the “three plots of land”, and the research on the adaptability of resource-constrained watersheds is insufficient [9]. In addition to this, insufficient attention has been paid to the synergy of the reform of the watershed scale, and food security research has formed a self-contained analytical framework. However, in the multi-dimensional framework of economy and policy, water shortage and the decline in the quality of arable land are the core constraints of grain production in the Yellow River Basin [10], and agricultural modernization is an important support [11]. The guaranteed assistance of policies such as grain production support and arable land protection has been verified [12]. However, the direct relationship mechanism between land system reform and food security still needs to be accurately empirically verified [13], and the existing food security assessment indicators are mostly applicable to the macro-scale. There is a lack of targeted tools to adapt to the county-level scale, and it is difficult to reflect micro-regional differences [14]; watershed scale research focuses on the Yangtze River, Black Water resource constraints, and arable land protection in river and other basins [15]. The research on the Yellow River Basin focuses on single issues such as arable land improvement and water resource allocation [16,17], and does not take land system reform as the core variable. Internationally, the research on the reform of land property rights in resource-constrained watersheds has confirmed the positive role of market-oriented allocation [18], but based on the private land property rights system, there are essential differences given the context of China’s collective land system, which makes international research difficult to directly learn from [19]. Agricultural modernization can be amplified through technology promotion and the popularization of agricultural machinery. The production effect of land reform [20,21], the impact of new urbanization on the effect of land reform, is different; some studies believe that it can reduce the cost of land transfer [22], and some studies warn of the risk of arable land encroachment [23]. This difference reflects the situational dependence of the regulation effect, but a study of the regulation mechanism under the triple constraint of “food–energy–ecology” in the Yellow River Basin remains absent. Based on the existing research, the academic community has accumulated rich achievements, but there are still three significant gaps. Firstly, from the perspective of research, there is a lack of special research on the relationship between the coordinated reform of the “three plots of land” in the Yellow River Basin and food security, and an evaluation of the effect of institutional reform on the basin scale has yet to be conducted [17]. Secondly, from a theoretical perspective, regarding systems, a complete analysis framework of the “land system reform–resource allocation optimization–food security improvement” relationship has not been built, and the regulatory role and boundary conditions of agricultural modernization and urbanization remain insufficiently explained. Thirdly, in terms of research scale, an accurate evaluation and spatial differentiation analysis of micro-county scales remains weak, making it difficult to support the formulation of a differentiated policy. Based on this, this study focuses on the Yellow River Basin, a typical resource-constrained basin, and deeply analyzes the impact of the “three plots of land” reform on food security and the adjustment mechanism, which has important academic value and practical significance.

Finally, the transformation from administrative allocation to market allocation must also be addressed; guided by the theory of sustainable food security, it tries to promote the development of agriculture in the Yellow River Basin in an efficient, green, and sustainable direction by improving the efficiency of land use and promoting the integration of agricultural modernization and new urbanization, so as to save grain on the land, while grain in the “technology” strategy provides institutional guarantee [24]. Judging by the practical results, by 2022, 6533.33 hectares of collective management construction land in the Yellow River Basin had entered the market, and 132,000 housing estates had been withdrawn for a fee, with direct economic benefits of RMB 21.8 billion (Yellow River Basin Reform Assessment Report, 2023). However, the effect of the reform policy shows significant spatial heterogeneity. Based on the cross-integration of land ownership theory, institutional change theory, and sustainable food security theory [25], this study focuses on the spatial effect adjustment mechanism of the “three plots of land” reform of the Yellow River Basin, and expands the specific policy evaluation to the general research category of resource-constrained watersheds. The core problem that this research seeks to address is the effect of the linear relationship and the heterogeneity mechanism on food security. The revised law of regional heterogeneity will also be addressed [26].

The core goals of this study are twofold. On the one hand, it seeks to clarify the adaptability of the “three plots of land” reform and the coordinated development of food security in different regions of the Yellow River Basin, and to provide micro-empirical support for the implementation of the strategy of “storing grain in the land and grain in technology” in the Yellow River Basin. On the other hand, it looks to refine the replicable experience of the reform of the land system in the resource-constrained basin, and to provide theoretical reference and practical plans for the coordinated promotion of global rural land system reform and food security, and finally to realize the dual improvement resulting from the practical significance of policies and academic contributions made on the basis of strengthening the scientific value of research.

2. Theoretical Analysis and Research Hypothesis

The “three plots of land” reform pilot programs were established in 2015. To this end, to maximize the synergistic and coordinated effects of policies and fully leverage the pilot’s impact on promoting food security, the reform adopts a three-pronged approach to maximize policy synergy [27]. The “three plots of land” reform effectively safeguards farmers’ income and property rights while driving land system transformation. It enhances food security development through new urbanization and strengthens county-level food security by advancing agricultural modernization. Based on this analysis, the research framework is presented in Figure 1, below.

2.1. Direct Impact of the “Three Plots of Land” Reform on Food Security

Rural land expropriation reform primarily focuses on standardizing expropriation procedures, raising compensation standards, and safeguarding the legitimate rights of affected farmers. On one hand, reasonable compensation ensures farmers receive adequate economic support after losing their land, maintaining their living standards and reducing social conflicts triggered by expropriation. This fosters a stable social environment conducive to food production. On the other hand, standardized expropriation procedures help ensure land is acquired for legitimate construction projects, preventing indiscriminate land seizures and resource wastage. This safeguards the relative stability of arable land area, providing fundamental land resource security for food safety. For instance, strictly defining the scope of public interest prevents the arbitrary expropriation of farmland for commercial development under the guise of public benefit, ensuring that high-quality arable land remains dedicated to food production.

The reform allows collectively owned operational construction land to enter the market, breaking the urban–rural land dual structure, enabling rural collective operational construction land to enter the market on equal footing with state-owned land, with equal rights and prices. This reform provides greater land factor support for rural industrial development [28]. On one hand, rural areas can utilize collectively owned operational construction land entering the market to develop industries such as agricultural product processing and rural tourism. This extends agricultural–industrial chains, enhances agricultural value-added, increases farmers’ income, and consequently boosts their motivation and capacity for grain production [29]. On the other hand, industrial development attracts capital, technology, and other factors to flow into rural areas, improving rural infrastructure and production conditions. This facilitates the promotion of advanced agricultural production techniques and management expertise, thereby enhancing grain production efficiency and quality.

Reforms to the homestead land system primarily involve improving the safeguarding of homestead land rights and acquisition methods, as well as exploring mechanisms for compensated use and withdrawal of homestead land rights [30]. By strengthening the protection of homestead land rights, farmers’ usufructuary rights over homestead land are clarified, stabilizing their residential expectations and enabling them to engage in agricultural production with peace of mind [31]. Exploring paid usage and exit mechanisms for homestead land can activate idle rural homesteads. The freed-up land resources can be converted into arable land through reclamation, thereby increasing cultivated land area. Simultaneously, the released homestead quotas can be utilized through policies like urban–rural construction land exchange to provide land quotas for urban development. This alleviates the conflict between urban expansion and arable land protection, indirectly safeguarding the cultivated land resources essential for grain production. Based on this, Hypothesis H1 is proposed.

H1: 

The “three plots of land” reform in the Yellow River Basin significantly promotes food security.

2.2. The Moderating Effect of Agricultural Modernization

Agricultural modernization encompasses multiple dimensions, including agricultural production, output, labor quality, and the agricultural ecological environment [32]. Against the backdrop of the “three plots of land” reform in the Yellow River Basin, agricultural modernization has driven the application of advanced technologies such as precision agriculture, biotechnology, and agricultural mechanization in grain production. The reform optimizes land resource allocation, creating conditions for the implementation of these technologies and helping to increase land productivity and grain output. Simultaneously, the promotion of agricultural mechanization enhances labor productivity and reduces labor intensity, making grain production more efficient. Its advantages are particularly pronounced in large-scale land transfers and concentrated operations. The land transfer measures within the “three plots of land” reform provide precisely the support needed for the large-scale application of mechanization [33]. Agricultural modernization drives the evolution of agricultural production organization toward diversification, scale expansion, and specialization [34]. During the “three plots of land” reform, these new agricultural business entities can better integrate resources such as land, capital, and technology to engage in large-scale, standardized grain production and management activities. Simultaneously, by collaborating with research institutions and universities, these new agricultural entities can more rapidly access advanced agricultural technologies and management expertise. This enhances the technological sophistication and management standards of grain production. Building upon the optimized allocation of land resources achieved through the “three plots of land” reform, they further improve the efficiency and quality of grain production, thereby safeguarding food security [35]. Agricultural modernization demands robust agricultural infrastructure. Driven by the rural economic development spurred by the “three plots of land” reform, investment in agricultural infrastructure has steadily increased. Well-developed irrigation systems ensure adequate water supply for crop growth, improve water resource utilization efficiency, and mitigate the impact of natural disasters like droughts on grain production [36]; convenient transportation facilities facilitate the movement of agricultural inputs and the marketing of grain products, lowering logistics costs; and advanced storage facilities reduce post-harvest losses and ensure grain quality [37]. Thus, agricultural modernization provides robust support for the “three plots of land” reform to enhance food security by improving agricultural infrastructure. Based on this, Hypothesis H2 is proposed.

H2: 

Agricultural modernization exerts a positive moderating effect on the relationship between the “three plots of land” reform and food security in the Yellow River Basin.

2.3. The Moderating Effect of New-Type Urbanization

The core of the new-type urbanization is “people as the core”, which forms an essential difference from the characteristics of traditional urbanization of “heavy on things and people, land expansion first”. This difference is the key premise for its ability to positively regulate the relationship between the “three plots of land” reform and food security. Traditional urbanization is guided by the expansion of urban scale. The large-scale expropriation of arable land and the one-way flow of rural resources into cities often leads to the fragmentation of arable land and the hollowing of rural areas. There is a contradiction between “people leaving the wasteland” and the coexistence of urban disorderly occupation, which poses a potential threat to food security [38]. New urbanization follows four principles: people-oriented, optimized layout, ecological civilization, and cultural inheritance [39]. The core characteristics are reflected in three aspects. The first is the civilization of the agricultural transfer population, through the reform of the household registration system and the equalization of basic public services, so that the transferred population can truly “live and integrate”. In this aspect, it is essential to reduce the dependence on land for survival. The second is the assurance of adherence to the integrated development of urban and rural areas, the breaking down of barriers to the flow of factors, and the promotion of the return of urban funds, technology, and talents to rural areas, rather than having a single “rural support city” [40]. The third is the intensive and sustainable use of land, adhering to the red line of arable land protection, and revitalizing land through stock, urban coordination of the planning of township land, and the optimization of the allocation of resources [41]. This essential difference makes new urbanization and food security form a “synergistic and mutual promotion” relationship, which can directly adjust the food security effect of the “three plots of land” reform. With the cooperation of the “three plots of land” reform, new urbanization has promoted the orderly transfer of the rural population. The decrease in rural population has relatively increased per capita land resources, creating conditions for moderate-scale operation and improving the efficiency of land use. For example, farmers who have settled in the middle reaches of the irrigation area of the Yellow River Basin have concentrated scattered plots in family farms, cooperatives, and other new agricultural business entities through land transfer, so as to realize the scale and intensification of grain production, and the scale efficiency has been significantly improved [42]. At the same time, the idleness of housing estates caused by the transfer of rural population (the idle rate of rural housing estates in the Yellow River Basin is about 12–18%), through reform measures such as the paid withdrawal of housing estates and the entry of collective management construction land into the market; the reclamation of land to supplement grain resources; and the development of supporting industries such as primary processing of agricultural products, continue to optimize land use. This is an example of an excellent approach that can be used [43]. New urbanization also drives the clustering and upgrading of urban secondary and tertiary industries. In the context of the “three plots of land” reform breaking the dual structure of urban and rural land, the connection between urban and rural industries and the flow of factors is strengthened. On the one hand, urban industries have changed from “scale expansion” to “quality improvement”, and the demand for agricultural products has changed from “quantity satisfaction” to “quality upgrading”, forcing rural areas to optimize their grain planting approach, improve the added value through in-depth processing [44], and motivating farmers to increase production input. On the other hand, the funds and technology remain gathered in urban areas. Technology, talents, and other factors have penetrated into rural grain production. For example, financial institutions have launched “grain production loans”, and scientific research institutions have promoted water-saving irrigation, soil measurement formula fertilization, and other technologies [45], combined with the large-scale scenarios formed by reform, quickly transforming into real production efficiency improvement. In summary, new urbanization has built a platform for the reform of “three plots of land” to optimize land allocation and activate the vitality of factors through the path of “population citizenship–two-way flow of factors–industrial upgrading”, significantly enhancing the role of reform in ensuring food security. Based on this, Hypothesis H3 is proposed.

H3: 

New-type urbanization plays a positive moderating role in the relationship between the “three plots of land” reform and food security in the Yellow River Basin.

2.4. Regional Heterogeneity Effects of the “Three Plots of Land” Reform on Food Security

In the upstream grain-producing regions of the Yellow River Basin (e.g., certain counties in Henan and Shandong), arable land resources are contiguous, soil fertility is relatively high, agricultural infrastructure is well-developed, and farmers possess strong grain-growing traditions and enthusiasm. Within the “three plots of land” reform, land expropriation reforms can more precisely protect high-quality farmland and prevent non-agricultural waste. The entry of collectively owned operational construction land into the market can efficiently meet industrial demands for agricultural product processing and storage, further extending the grain supply chain. Idle land revitalized through homestead reforms can also be more readily converted into effective farmland—aligning closely with the functional positioning of the midstream “core grain production zone”. Policy dividends can directly and efficiently contribute to food security, making this reform’s positive impact on food security the strongest [46]. Although midstream regions, such as parts of Dongying in Shandong Province near the Yellow River estuary, possess some arable land resources, they face potential pressure from non-agricultural land use due to rapid urbanization and intensified competition between industrial and agricultural land. While the “three plots of land” reform can safeguard arable land quantity by standardizing land acquisition and revitalizing land resources and enhance grain production returns by entering the market to develop supporting agricultural industries [47], compared to the midstream, some counties in the downstream region exhibit a slightly lower priority for grain production within regional development due to their diversified economic structures. Consequently, the reform’s ability to strengthen food security is somewhat weaker, resulting in a secondary effect. Primarily located on the Qinghai–Tibet Plateau and western Loess Plateau (e.g., parts of Qinghai’s Haidong and Gansu’s Linxia counties), these areas face constraints from high altitude, arid conditions, and limited rainfall. Their arable land is fragmented, with low fertility; agricultural infrastructure is weak; and farmers’ grain-growing profits remain relatively limited. Within the “three plots of land” reform, revitalized homestead land faces significant challenges in reclamation and low efficiency in converting to effective farmland. The entry of collectively owned operational construction land into the market struggles to form grain industry chains due to weak industrial foundations. Land expropriation reforms also see limited practical effects on farmland protection constrained by natural conditions; the reform’s alignment with the upstream region’s characteristics of “ecological conservation priority and constrained grain production conditions” is relatively low, preventing the full realization of policy dividends. Consequently, these reforms exhibit the weakest positive impact on food security. Based on this, Hypothesis H4 is proposed.

H4: 

The policy empowerment effects of the “three types of land” reforms in the Yellow River Basin on food security exhibit significant spatial differentiation, with an overall gradient pattern of “midstream > downstream > upstream.”

3. Materials and Methods

3.1. Data

The reform pilot of “three plots of land” is mainly concentrated in the county area, so all the raw data in this article come from the Chinese County Statistical Yearbook. For the exploration of the impact of the “three plots of land” reform in rural areas on the food security of counties, the final sample interval selected in this study is from 2010 to 2023. There are two reasons for the selection of the time interval. Firstly, the DID model requires that samples be kept before and after the implementation of the policy, and secondly, although the pilot policy of “three plots of land” reform ended at the end of 2019, “prudent promotion” is the consensus and tone of the reform of the rural land system. After the pilot, the reform was still not rapid at the national level. In order to go beyond this, the research period was extended to be up until 2023. In terms of sample selection, most of the experimental design groups and control groups evaluated the policy effect [48,49]. Drawing on this research form, this article first divided the sample area into experimental groups and control groups. The scope of the experimental group is the “On Authorized Countries of the Standing Committee of the National People’s Congress”; from the list of pilot areas determined by the Academy in the Decision on the Temporary Adjustment and Implementation of Relevant Legal Provisions in the Administrative Areas of 33 Pilot Counties (Cities and Districts), such as Daxing District, Beijing prevailed. Since the area studied is the Yellow River Basin, after confirming the scope of 11 experimental groups in the pilot counties (cities and districts), according to the geographical location, the groups were defined as either agricultural or economic. The structure is similar to the principle of development level, and the areas from the pilot counties and cities that were used in the pilot of the reform were selected as the control group. This article eliminates the observation groups with serious missing variable values. Some of the missing data are supplemented by the linear interpolation method to standardize the data. This paper finally adopts more than 1000 panel data from the county.

3.2. Variable Selection

3.2.1. Core Explanatory Variable

The “three plots of land” reform constitutes the core component of rural land system reform. This study employs the interaction term between the dummy policy variable for the “three plots of land” reform pilot and the dummy time variable as the core explanatory variable. For the dummy policy variable, counties not participating in the “three plots of land” reform pilot serve as the control group and are assigned a value of 0, while those participating in the pilot serve as the experimental group and are assigned a value of 1. For the dummy time variable, the start year of the pilot program serves as the policy shock point. Values are 0 for years before 2015 and 1 for 2015 and later. By constructing the interaction term between these two variables and estimating its coefficient using a difference-in-differences model, we derive the net effect of the pilot policy.

3.2.2. Explained Variable

Based on the Food and Agriculture Organization of the United Nations’ definition of food security and China’s specific circumstances as a food-producing nation, this study adopts the entropy method proposed by Zhu et al. to construct the following food security evaluation index system [50,51], as shown in

Table 1. Food security evaluation indicator system.

Primary Indicators	Secondary Indicators	Variable Explanation	Indicator Attribute
Food supply security	Grain Sown Area	County grain sown area (million hectares)	+
	Total Agricultural Machinery Power	Total machinery power per unit of cultivated area (kilowatts/hectare)	+
	Local government general budget expenditures	Total budget amount of various expenditure items (RMB 10,000)	+
Food access security	Per capita grain availability	Ratio of total grain output to resident population (tons/person)	+
	Rural per capita disposable income	Rural per capita disposable income (RMB 10,000)	+
	Rural Engel’s Coefficient	Proportion of rural food consumption expenditure to total consumption expenditure	−
Stability and security of food production	Grain yield volatility rate	(Current year’s total grain output—Average of total grain output over the past five years)/Current year’s total grain output	−
	Grain price volatility rate	(Current year’s grain price index—Previous year’s grain price index)/Current year’s grain price index	−
	Affected area of crops	Affected area of crops/Sown area of crops	−
Environmental sustainability of food production	Pesticide usage	Pesticide application per unit of grain-sown area (ton/hectare)	−
	Chemical fertilizer usage	Chemical fertilizer application per unit of grain sown area (ton/hectare)	−
	Plastic film usage	Plastic film application per unit of grain-sown area (ton/hectare)	−

Note:“+”means that the higher the index value, the better the performance of the evaluation object. “−” means that the higher the index value, the worse the performance of the evaluation object.

. It comprises four primary indicators—supply, access, stability, and sustainability—along with ten secondary indicators. The rationale for selecting these indicators is as follows. Firstly, regarding supply, excluding the impact of international agricultural trade on food security, the core issue of food security is grain supply. The supply dimension comprises three secondary indicators: “grain sown area”, “total agricultural machinery power”, and “local government general budget expenditure”. Secondly, accessibility encompasses both the physical availability of food at the household level and its economic affordability at the family level. Accessibility is measured using “per capita grain availability”, “rural per capita disposable income”, and “rural Engel’s coefficient”. The third metric is stability. Maintaining stable food production and effective supply is key to enhancing food supply chain resilience. This is measured by three secondary indicators: “food production volatility”, “food price volatility”, and “crop disaster-affected area”. The fourth metric is sustainability. The environmental sustainability of grain production is crucial for addressing extreme agricultural non-point source pollution, deteriorating ecosystems, and ensuring food security. This is measured by three secondary indicators: “pesticide application volume”, “fertilizer application volume”, and “plastic film usage volume”. The intensive application of fertilizers and pesticides is unsustainable for increasing farmland yields, leading to soil compaction and pollution. Similarly, the use of plastic film causes pollution, undermining the long-term stability of grain production. All three indicators are negative indicators.

• Non-quantification processing;

In the comprehensive evaluation, due to the different units of measurement and types of each evaluation indicator, in order to exclude the impact of these differences, it is necessary to carry out a non-quantitative treatment of these evaluation indicators to make them comparable, and this can also avoid the appearance of extreme values in sample data. Commonly used dimensionless processing methods include the standardized processing method, extreme value processing method, linear proportional method, normalized processing method, vector normalization method, and efficiency coefficient method. Because the extreme value processing method has the advantages of monotony, difference ratio invariance, translation independence, scaling independence, interval stability, etc., the extreme value method is used for dimensionless processing of the original data. For those positive indicators, that is, where a larger indicator value means it is more beneficial to the development of the system, Formula (1) is adopted; for those negative indicators, that is, where a smaller indicator value means that it is more beneficial to the development of the system, Formula (2) is adopted, and for the processing of moderate indicators, Formula (3) is adopted.

$${\mathrm{X}}_{\mathrm{i}\mathrm{j}}^{″}={\displaystyle \frac{{\mathrm{X}}_{ij}-min\left(X_j\right)}{max\left(X_j\right)-min\left(X_j\right)}}$$

(1)

$${\mathrm{X}}_{\mathrm{i}\mathrm{j}}^{″}={\displaystyle \frac{max\left(X_j\right)-X_{ij}}{max\left(X_j\right)-min\left(X_j\right)}}$$

(2)

$${\mathrm{X}}_{\mathrm{i}\mathrm{j}}^{″}={\displaystyle \frac{|{\mathrm{X}}_{\mathrm{i}\mathrm{j}}-\frac{\mathrm{M}+\mathrm{m}}{2}|}{\mathrm{m}\mathrm{a}\mathrm{x}\left\{{\mathrm{X}}_{\mathrm{j}}\right\}-\mathrm{m}\mathrm{i}\mathrm{n}\left\{{\mathrm{X}}_{\mathrm{j}}\right\}}}\text{,}\mathrm{Among}\mathrm{M}=\mathrm{M}\mathrm{a}\mathrm{x}\left\{{\mathrm{X}}_{\mathrm{j}}\right\},\mathrm{m}=\mathrm{m}\mathrm{i}\mathrm{n}\left\{{\mathrm{X}}_{\mathrm{j}}\right\}$$

(3)

• Data normalization processing

Normalization can obtain the probability of the observation of a sample in an indicator subsystem, for which it is convenient to use the information entropy formula to measure the uncertainty of the indicator subsystem. The specific processing method is as shown in Formula (4).

$$Y_{ij}^{\prime }={\displaystyle \frac{X_{ij}^{\prime }}{\sum _{i=1}^m{Y}_{ij}}}$$

(4)

• Calculate the entropy of relative information;

$$e_j=-k\sum _{i=1}^m{Y}_{ij}\times \ln Y_{ij}),k={\displaystyle \frac{1}{\ln \mathrm{m}}},0\le e\le 1$$

(5)

• Calculate the redundancy and weight of information entropy

$$d_j=1-e_j$$

(6)

$$W_j={\displaystyle \frac{d_j}{\sum _{j=1}^n{d}_j}}$$

(7)

• Calculate the comprehensive score

$$S_i=\sum w_j\times X_{ij}^{\prime }$$

(8)

New-type urbanization and agricultural modernization are calculated by this method.

3.2.3. Control Variables

Given the limited availability of county-level data, this study selected control variables spanning economic/industrial, social, and demographic dimensions to mitigate endogeneity issues from omitted variables. Key variables include economic development level, industrialization level, social welfare, agricultural development level, agricultural technology, and population size. Descriptive statistics for relevant variables are presented in

Table 2. Variable definitions.

Variable Type	Variable Name	Variable Definition and Unit
Explained variable	FS	Calculated using the entropy method
Core Explanatory Variable	DID	DID interaction term (post×treat)
Moderating variables	NTU	Calculated by the entropy method
	AML	Calculated by the entropy method
Control Variables	CAA	Crop area affected by natural disasters (e.g., drought, flood, windstorm, etc.) (hectares)
	TSA	Total sown area of all crops (hectares)
	AFAFE	Number of laborers engaged in agriculture, forestry, animal husbandry, fishery production, and related services (persons)
	TAMP	Total power of various machinery used for agricultural production (10,000 kWs)
	GGBR	Income obtained for local fiscal expenditure (RMB)
	CCA	Land area available for normal cultivation within the specified year (hectares)
	PIVA	Economic value created through primary industry production activities such as agriculture, forestry, animal husbandry, and fishery (RMB)

. To enhance the estimation efficiency of the model, the following control variables were incorporated into the equations: affected crop area (CAA), total sown area (TSA); agricultural, forestry, animal husbandry and fishery employees (AFAFE); total agricultural machinery power (TAMP); government general budget revenue (GGBR); common cultivated area (CCA); and primary industry value added (PIVA) [52].

3.2.4. Adjustment Variable

The moderating variable in this study is the level of new urbanization, calculated using the entropy method. Following principles of scientific rigor, representativeness, and data availability, this paper constructs a comprehensive evaluation index system for new-type urbanization’s productive capacity. Drawing on existing index systems and referencing official new-type urbanization planning documents, this study defines new urbanization as the coordinated development of three subsystems: population, economy, and society. Eleven indicators were selected to measure new-type urbanization (

Table 3. New-type urbanization evaluation indicator system.

Primary Indicator Level	Second-Level Indicator Layer	Unit	Direction
Population Urbanization	Urbanization rate of resident population	%	+
	Number of students enrolled in regular institutions of higher education	10,000 persons	+
	Number of employed workers in urban units	10,000 persons	+
Economic urbanization	GDP per capita	RMB/person	+
	Added value of the secondary industry	RMB 100 million	+
	Added value of the tertiary industry	RMB 100 million	+
	Export volume	Thousand dollars	+
Social Urbanization	Number of medical institution beds per 10,000 people	Beds/10,000 people	+
	Number of various social welfare institutions	Units	+
	Total electricity consumption of society	10,000 kWh	+
	Urban–rural income gap	RMB	−

Note:“+” means that the higher the index value, the better the performance of the evaluation object. “−” means that the higher the index value, the worse the performance of the evaluation object.

) [53]. Regarding the selection of agricultural modernization evaluation indicators, scholars predominantly establish evaluation frameworks centered on agricultural input and output segments. Drawing from the indicator selection by Li Zhanwen et al. and grounded in fundamental principles of availability, precision, and objectivity [54], four primary indicators and ten secondary indicators are adopted as agricultural modernization evaluation metrics (

Table 4. Agricultural modernization evaluation indicator system.

Primary Indicator Level	Second-Level Indicator Layer	Variable	Direction
Agricultural production	Level of mechanization	Total power of agricultural mechanization (kW/hectare)	+
	Level of technological advancement	Area of facility agriculture (hectares)	+
	Scale level	Mechanically harvested area (hectares)	+
Agricultural output	Gross output value of agriculture, forestry, animal husbandry, and fishery	Gross output value of agriculture, forestry, animal husbandry, and fishery (RMB 10,000)	+
	Level of grain output	Total grain output (tons)	+
	Farmer income level	Disposable income of farmers (RMB 10,000)	+
Quality of labor force	Educational level of employees	Number of rural employees (persons)	+
	Cultivated land area per laborer	Commonly used cultivated land area (hectares)	+
Agricultural ecological environment	Chemical fertilizer usage	Fertilizer usage (tons/hectare)	−
	Agricultural plastic usage	Agricultural plastic film usage (tons/hectare)	−

Note: “+”means that the higher the index value, the better the performance of the evaluation object. “−” means that the higher the index value, the worse the performance of the evaluation object.

). Among these, fertilizer usage and plastic film usage serve as negative indicators.

3.3. Model Construction

3.3.1. Difference-in-Differences Model

China’s vast territory and large population exhibit significant regional disparities in natural resource endowments and socioeconomic development levels. These baseline differences generate temporal effects through inertia and trends over time, complicating research on the causal relationship between rural land system reforms and food security. In 2015, China launched pilot reforms for the “three types of land”. The variation in policy impacts across counties provided a quasi-natural experiment for identifying the treatment effects of rural land system reforms. To empirically examine whether these reforms affect county-level food security, this study utilizes this quasi-natural experiment to construct a difference-in-differences model. The specific baseline model is as follows:

$$f{s}_{it}={\beta }_0+\beta Polic{y}_{it}+\alpha Control{s}_{it}+{\mu }_i+{\gamma }_t+{\xi }_{it}$$

(9)

In Equation (9), i denotes region, t denotes year, represents the food security level of county i in year t, $Polic{y}_{it}$ is the core explanatory variable indicating the implementation status of the “three plots of land” reform in region i during year t, $Control{s}_{it}$ denotes a set of control variables,${\mu }_i$ represents region fixed effects, ${\gamma }_t$ denotes year fixed effects, and ${\xi }_{it}$ represents the random disturbance term [55].

3.3.2. Moderated Effects Model

To verify the impact of new-type urbanization and agricultural modernization on food security development under the “three plots of land” reform policy, models (10) and (11) were constructed based on model (9) by incorporating new-type urbanization and agricultural modernization. These models examine their moderating effects on the relationship between the “three plots of land” reform and food security.

$$f{s}_{it}={\beta }_0+{\beta }_{1}Polic{y}_{it}+{\beta }_{2}NT{U}_{it}+{\beta }_{3}Control{s}_{it}+{\mu }_i+{\gamma }_t+{\xi }_{it}$$

(10)

$$f{s}_{it}={\beta }_0+{\beta }_{1}Polic{y}_{it}+{\beta }_2{\mathrm{A}\mathrm{M}\mathrm{L}}_{it}+{\beta }_{3}Control{s}_{it}+{\mu }_i+{\gamma }_t+{\xi }_{it}$$

(11)

Here, (Agricultural Modernization Level) AML denotes the level of agricultural modernization, and (New-Type Urbanization) NTU denotes the level of new-type urbanization

4. Empirical Results

4.1. Descriptive Statistical Analysis

Following data collection, descriptive statistical analysis was first conducted, as shown in

Table 5. Descriptive statistics for each variable.

Variable	Obs	Mean	Std. Dev.	Min	Max
FS	1530	0.336	0.077	0.13	0.574
DID	1530	0.052	0.223	0	1
NTU	1530	0.623	0.063	0.133	0.696
AML	1530	0.177	0.096	0.009	0.435
CAA	1530	6.775	0.74	1.099	8.075
TSA	1530	3.966	0.637	1.946	5.613
AFAFE	1530	11.428	0.723	9.178	13.165
TAMP	1530	3.548	0.825	1.099	5.645
GGBR	1530	10.797	1.121	7.279	14.6
CCA	1530	10.721	0.616	8.807	12.223
PIVA	1530	11.972	0.828	9.304	13.706

, presenting the trends in food security and related indicators across county-level regions in the Yellow River Basin before and after the reform. Preliminary trends indicate that food security in most counties affected by the reform showed a steady upward trajectory within the five years following the reform.

4.2. Regression Analysis Results

Table 6. Benchmark regression results.

	(1)	(2)	(3)	(4)	(5)	(6)	(7)
	FS	FS	FS	FS	FS	FS	FS
DID	0.0237 *** (0.0031)	0.0237 *** (0.0031)	0.0182 *** (0.0030)	0.0188 *** (0.0030)	0.0085 *** (0.0025)	0.0078 *** (0.0025)	0.0058 ** (0.0024)
CAA	−0.0041 *** (0.0007)	−0.0041 *** (0.0007)	−0.0038 *** (0.0007)	−0.0037 *** (0.0006)	−0.0032 *** (0.0005)	−0.0032 *** (0.0005)	−0.0033 *** (0.0005)
TSA		0.0089 (0.0074)	0.0065 (0.0071)	0.0043 (0.0070)	−0.0002 (0.0059)	−0.0005 (0.0059)	−0.0029 (0.0056)
AFAFE			−0.0557 *** (0.0046)	−0.0518 *** (0.0047)	−0.0125 *** (0.0043)	−0.0131 *** (0.0043)	−0.0078 * (0.0041)
TAMP				0.0077 *** (0.0020)	0.0050 *** (0.0017)	0.0049 *** (0.0017)	0.0066 *** (0.0016)
GGBR					0.0182 *** (0.0007)	0.0177 *** (0.0008)	0.0079 *** (0.0011)
CCA						0.0090 ** (0.0039)	0.0028 (0.0037)
PIVA							0.0236 *** (0.0019)
Fixed Effects	Yes	Yes	Yes	Yes	Yes	Yes	Yes
_cons	0.3630 *** (0.0047)	0.3278 *** (0.0297)	0.9715 *** (0.0607)	0.9078 *** (0.0627)	0.2878 *** (0.0582)	0.2041 *** (0.0685)	0.0373 (0.0666)
N	1530	1530	1530	1530	1530	1530	1530
R2	0.061	0.062	0.148	0.157	0.409	0.411	0.466

Note: Numbers in parentheses indicate standard errors; “***”, “**”, and “*” denote significance at the 1%, 5%, and 10% levels, respectively.

reports the benchmark regression results, indicating the impact of the “three plots of land” reform in the Yellow River Basin on food security. treated_t ∗ post_t is the DID variable, whose coefficient represents the effect of the “three plots of land” reform policy on food security in the Yellow River Basin. Control variables include affected crop area (CAA); total sown area (TSA); agricultural, forestry, animal husbandry and fishery employees (AFAFE); total agricultural machinery power (TAMP); government general budget revenue (GGBR); common cultivated area (CCA); and primary industry value added (PIVA).

The results in Column (1) indicate that the “three plots of land” reform significantly enhanced food security levels. To control for other factors, control variables were progressively added to Column (1); Column (7) shows that the DID coefficient remains significantly positive even after incorporating all control variables, indicating robustness of the results. This confirms that the “three plots of land” reform policy substantially improved food security levels, and that this effect would be underestimated without controlling for these variables. The benchmark regression results demonstrate that the “three plots of land” reform effectively enhanced food security. Hypothesis H1 has therefore been verified. To verify the robustness of its impact on food security, a robustness regression is conducted below.

4.3. Parallel Trends Test

A key assumption of the DID model is satisfying the parallel trends test. This study employs the event study method proposed by Jacobson et al. [56] to conduct the parallel trends test, using the year the “three plots of land” reform pilot program commenced as the baseline year. The parallel trend test results are shown in Figure 2. Pilot counties and non-pilot counties exhibited no significant differences prior to policy implementation but showed significant divergence afterward, satisfying the parallel trend test assumption [57].

4.4. Placebo Test

To assess the extent to which regression results are affected by omitted variables and random factors, individual and time-specific regressions were randomly generated for food security counties [58,59], repeated 500 times, and the resulting distribution of estimated coefficients is shown in Figure 3. As seen in Figure 3, the estimated coefficients are primarily clustered around zero, indicating no severe omitted variable issues and robust conclusions.

4.5. Moderating Effect Analysis

The moderating effect of new-type urbanization is shown in Column (1) of

Table 7. Analysis of regulatory effect results.

Variable	(1)	(2)
DID	0.010 * (1.654)	0.018 *** (4.592)
NTU	0.134 ** (2.231)
AML		−0.018 *** (−4.628)
fs × NTU	0.416 ** (2.329)
fs × AML		0.027 * (1.659)
_cons	0.331 *** (187.810)	0.339 *** (352.547)
Fixed Effects	yes	yes
N	1530	1530
R2	−0.0228	0.9388

Note: Numbers in parentheses indicate standard errors; “***”, “**”, and “*” denote significance at the 1%, 5%, and 10% levels, respectively.

. The coefficient value for fs × NTU is 0.416 and significant at the 5% level. This indicates that new-type urbanization exerts a positive moderating effect on the influence of the “three plots of land” reform policy on food security, validating Hypothesis H3 [60]. The moderating effect of agricultural modernization is shown in Column (2) of Table 7. The coefficient value of fs × AML is 0.027 and is significant at the 10% level, indicating that agricultural modernization exerts a positive moderating effect on the process by which the “three plots of land” reform policy impacts food security. Hypothesis H2 is thus validated.

4.6. Regional Heterogeneity Analysis

Against the backdrop of implementing the “three plots of land” reform in the Yellow River Basin, policy effects did not manifest uniformly across all regions. Considering the differences in geographical location, climatic conditions, economic development levels, and social structures among various counties in the basin, the policy implementation exhibited significant regional heterogeneity. Therefore, regional heterogeneity analysis was incorporated into this study to explore in depth the response mechanisms and variations to the “three plots of land” policy across different areas.

County-level areas in the Yellow River Basin can be broadly divided into three regions: upstream, middle, and downstream. The upstream region encompasses Qinghai and Gansu, where the Yellow River originates. These areas face severe constraints in agricultural production due to geographical limitations and relatively scarce land resources. The middle region includes Shaanxi and Shanxi, characterized by abundant arable land, a stronger economic foundation, and significant influence from climatic conditions. The downstream region comprises Shandong and Henan, distinguished by favorable agricultural conditions and relatively advanced economies, serving as the primary grain-producing areas of the basin.

Through a Double Difference (DID) analysis of panel data from these two major regions, the study revealed spatial gradient differences in policy effectiveness, as shown in

Table 8. Regional heterogeneity test results.

	(1)	(2)	(3)
	Upstream	Midstream	Downstream
DID	0.0037 (0.0026)	0.0164 *** (0.0082)	0.0018 ** (0.0047)
CAA	−0.0035 *** (0.0006)	−0.0026 * (0.0015)	−0.0038 *** (0.0010)
TSA	−0.0010 (0.0050)	−0.0742 * (0.0379)	0.0307 (0.0247)
AFAFE	−0.0141 *** (0.0040)	−0.0040 (0.0177)	0.0140 (0.0114)
TAMP	0.0077 *** (0.0025)	0.0093 *** (0.0030)	0.0303 *** (0.0065)
GGBR	0.0053 *** (0.0013)	0.0007 (0.0033)	0.0222 *** (0.0022)
TCCA	0.0085 ** (0.0036)	−0.0027 (0.0137)	−0.0693 ** (0.0317)
PIVA	0.0209 *** (0.0021)	0.0375 *** (0.0052)	0.0034 (0.0050)
Fixed Effects	Yes	Yes	Yes
_cons	0.0965 (0.0640)	0.2374 (0.2589)	0.3865 (0.4203)
N	870	330	330
adj. R2	0.485	0.316	0.612

Note: Numbers in parentheses indicate standard errors; “***”, “**”, and “*” denote significance at the 1%, 5%, and 10% levels, respectively.

. The results indicate that policy effects exhibit a spatial gradient pattern of midstream > downstream > upstream.

The upstream region faces significant agricultural production challenges due to geographical constraints and climatic conditions. Scarce land resources further complicate the implementation of the “three plots of land” policy. While policy reforms have promoted optimized land resource allocation, the region’s grain production levels struggle to improve significantly due to inherent natural limitations. Consequently, policy effectiveness is relatively weak here. Land transfer activity among farmers in upstream regions remains low, with limited implementation of collective land market entry and homestead land withdrawal, resulting in low levels of intensive land use. Policy effects in midstream regions significantly outperform those in upstream areas, primarily due to the region’s relatively developed agricultural foundation and abundant land resources. With the advancement of the “three plots of land” policy, land transfer activity in the middle region has increased. The expropriation of rural land and the entry of collectively owned construction land into the market have become feasible, effectively promoting the scaling and modernization of agricultural production. Furthermore, the integration of policy implementation with new urbanization and agricultural modernization has further strengthened the midstream region’s capacity for food security. Consequently, food security in the midstream region has improved to a certain extent following policy implementation, with agricultural production models gradually shifting toward greater efficiency and intensification. With relatively abundant land resources and better infrastructure, the downstream area has become an area displaying remarkable results of the “three plots of land” reform policy. The region’s land transfer market is relatively mature, and the implementation of policies allowing collective land to enter the market and the withdrawal of homestead land has strongly promoted intensive land use. Simultaneously, the downstream region’s rapid new urbanization process and higher level of agricultural modernization provide robust support for increased grain production. Consequently, the “three plots of land” policy in the downstream region has a significant impact on food security, enhancing the stability and assurance capacity of grain production.

Regional variations in policy outcomes stem not only from geographical and climatic conditions but also from local economic development levels and governance capabilities. Upstream regions, constrained by scarce land resources and weak agricultural foundations, face greater implementation challenges. In contrast, midstream and downstream areas benefit from more abundant land resources and higher economic development levels, yielding more pronounced policy effects. The introduction of new-type urbanization and agricultural modernization as moderating variables has amplified their role in safeguarding food security in mid- and downstream regions, while their impact remains relatively limited in upstream areas.

Through the above regional heterogeneity analysis, this study further reveals the implementation effects of the “three plots of land” across different regions and the underlying mechanisms. Hypothesis H4 is validated. For advancing the policy in the Yellow River Basin, differentiated policy measures tailored to the actual conditions of each region are necessary to achieve sustainable agricultural production and effectively safeguard food security.

5. Discussion

5.1. The Positive Impact of the “Three Plots of Land” Reform on Food Security and Research Comparison

Based on the panel data of 101 county-level cities in the Yellow River Basin from 2010 to 2023, this study confirms that the “three plots of land” reform has had a significant positive impact on food security by optimizing the allocation of land resources. Its core pathways are reflected in the circulation of contracted land to promote large-scale agricultural operation, the paid withdrawal of housing estates to revitalize idle arable land resources, and the entry of collective operational construction land into the market to raise investment funds for agricultural production. The three work together to improve agricultural production efficiency and strengthen the capacity of food safety and security. This conclusion echoes similar studies in China. Song (2024) verified that the clarity of land property rights and the reform of the transfer system can reduce transaction costs and significantly improve the efficiency of land use through the study of the process of land circulation in rural areas in China [61]; the empirical results of Zhang and others (2021) also show that orderly farmland transfer can increase grain yield by 8–10% through large-scale operation [62]. Compared with international related studies, the conclusion of this study not only fits the core logic of global land system reform, but also highlights the uniqueness of China’s situation. The joint report of the Food and Agriculture Organization of the United Nations and the World Food Program (2021) pointed out that the optimal allocation of land resources is the key path to improving food security in resource-constrained regions, which is consistent with the core findings of this study; Johnson S R and others’ research on the market-oriented reform of agricultural land in Ukraine also confirmed that the linkage of large-scale land operation and capital investment can significantly improve the efficiency of grain production [63]. However, unlike the international single farmland reform based on private land property rights, China’s “three plots of land” reform takes collective land property rights as the institutional basis, builds a framework for “three plots of land” coordinated reform, accurately adapts to the special situation of “grain–energy–ecology” triple constraints in the Yellow River Basin, producing an alternative to single-land-type reform. The gap in revolutionary research has provided a new practical paradigm for the security of food security in countries with global collective land systems.

5.2. Analysis of the Positive Adjustment Role and Mechanism of New-Type Urbanization and Agricultural Modernization

The study found that new urbanization and agricultural modernization played a significant positive regulating role in the process of the China’s “three plots of land” reform, which takes collective land property rights as the institutional basis, builds a framework for “three plots of land” coordinated reform, affecting food security, and this regulatory effect is more prominent in the middle and downstream areas. This result has been further confirmed by comparison with domestic and foreign studies. At the level of agricultural modernization, agricultural production, agricultural output, worker quality, agricultural ecological environment, and other factors have significantly improved the adaptation efficiency of land resource allocation and grain production. This is consistent with the research conclusion of Yao Shi et al. (2023) on resource-constrained watersheds; that is, agricultural technological progress can amplify the production empowerment effect of land system reform [64]. Internationally, the research of Tesfaye B et al. (2023) on Ethiopia also confirms that the reform of land confirmation needs to be promoted in conjunction with the promotion of agricultural technology in order to effectively translate into an increase in grain production, which further highlights the key value of agricultural modernization as a ”reform effect amplifier” [65]. At the level of new urbanization, population concentration and the improvement of urban and rural infrastructure have reduced the transaction cost of land transfer and agricultural production. In counties where the urban population accounts for more than 50%, the reform effect has increased by 19.7%, while the county, which accounts for less than 30%, is only 8.3%. This discovery is different from the warnings of some international studies; Li and others (2025) once proposed that urbanization may lead to the encroachment of arable land and weaken the food security effect of land reform. However, the Yellow River Basin strictly implements the “red line of arable land protection”, and through the intensive use of urban land and the overall planning of urban and rural land, the coordinated development of towns and food security is achieved [66]. This differentiated result shows that there are boundary conditions for the adjustment of urbanization to the effect of land reform. Only urbanization based on the protection of arable land and the coordinated development of urban and rural areas can effectively support the improvement of food security.

5.3. Regional Heterogeneity Characteristics and Causes and Research Inspiration

The impact of the China’s “three plots of land” reform takes collective land property rights as the institutional basis, builds a framework for “three plots of land” coordinated reform, and the effect of this reform on food security presents a spatial gradient of “midstream > downstream > upstream”. The formation of this regional heterogeneous feature is the result of the joint action of multiple factors such as geographical location, climatic conditions, economic development level, and agricultural development mode, and echoes the core views of relevant research at home and abroad. From the perspective of the cause, the midstream area is dominated by irrigation areas, with sufficient water resources and an excellent quality of arable land, while the agricultural model is dominated by grain production. The basic conditions of land transfer and large-scale operation are superior, and it has become the area with the most significant reform effect; the downstream areas have a high degree of marketization, and the funds for collective management construction land entering the market. The benefits are more abundant, but the competition between industry, urban construction, and agricultural land is relatively fierce, which has diluted the focus effect of reform on grain production to a certain extent; the upstream areas are located in arid and semi-arid areas, the fragmentation of arable land is serious, the agricultural infrastructure is weak, and agricultural production is mainly small-scale survival agriculture, which limits the ability of the reform to optimize the space for resource allocation. This is consistent with the conclusion of Mutabazi’s research on Rwanda; that is, natural endowment and economic foundation are the key variables that determine the regional effect of land reform [67]. The transnational comparative study of Mabvundwi M and others also pointed out that the success of land system reform is highly dependent on local situational adaptation [68]. This heterogeneous characteristic provides a clear direction for policy optimization; the midstream areas should continue to deepen the integration of large-scale operation of contracted land and agricultural mechanization, and further release the scale effect, while downstream areas need to balance the relationship between industrial and agricultural land, guide the income of collective management construction land into the market to grain production, and develop a high-quality grain industry. Upstream areas should combine reform with the construction of water-saving irrigation facilities and ecological protection, and solve the problem of arable land fragmentation through small-scale farmland water conservancy construction and a cooperative model. At the theoretical level, this study enriches the analytical framework of “institutional reform→resource allocation→food security” by revealing the mechanism of regional heterogeneity, which provides an important reference for the formulation of differentiated land policies and the balance of food security and regional development in global resource-constrained regions.

6. Conclusions and Policy Recommendations

6.1. Conclusions

This study analyzes panel data from 101 county-level cities in the Yellow River Basin between 2010 and 2023 to evaluate the impact of the “three plots of land” reform policy on food security. It further examines the moderating effects of new urbanization and agricultural modernization on policy outcomes. The key findings are as follows. (1) The “three plots of land” reform policy has a significant positive impact on food security; by optimizing land resource allocation, the policy promotes rural land transfers, the entry of collectively owned construction land into the market, and the withdrawal of homestead land. This enhances agricultural production efficiency and consequently strengthens food security safeguards. (2) New-type urbanization and agricultural modernization exert positive moderating effects; advancements in these areas significantly amplify the implementation outcomes of the “three plots of land” policy, with particularly pronounced impacts in mid- and downstream regions. (3) Regional heterogeneity mechanisms are clearly defined; responses to the “three plots of land” reform policy vary significantly across different regions of the Yellow River Basin based on geographical location, climatic conditions, and economic development levels. Policy outcomes exhibit a spatial gradient of “midstream > downstream > upstream”. Additionally, differing agricultural development models and land use patterns across regions lead to variations in the policy’s effectiveness for food security implementation.

6.2. Policy Recommendations

Based on research findings regarding the positive impact of the “three plots of land” reform policy on food security, showing its positive regulatory role in new-type urbanization and agricultural modernization, and regional heterogeneity characteristics, the following three targeted policy recommendations are proposed to further enhance reform effectiveness, safeguard food security, and promote coordinated regional development:

6.2.1. Implement Differentiated “Three Plots of Land” Reform Strategies to Precisely Align with Regional Development Needs

It is essential to follow the differences in the coefficients of policy effects in each region, abandon the “one-size-fits-all” model, and match regional resource endowment and reform needs in a targeted manner. In the midstream areas (the policy effect is the strongest, the coefficient is 0.0164), we will focus on “unleashing large-scale effects” and deepen reform coordination. Firstly, the restrictions on the scale of collective management construction land entering the market must be relaxed, and more than 30% of the income from entering the market should be used for high-standard farmland construction. A greater concentration of contracted land should be afforded to large grain growers, and a continuous grain production base should be built. Secondly, a supplementary linkage mechanism between housing base withdrawal and arable land should be established, and the reclamation of idle housing estates for arable land should take place. A subsidy of RMB 30,000–40,000 per hectare should be allotted for the production of wheat, corn, and other staple grains. Thirdly, the filing process of contracted land transfer should be simplified, the “online circulation + offline verification” mode should be implemented, the transfer cost should be reduced, and there should be an aim to increase the land circulation rate in the midstream area to more than 45% (currently about 38%). Downstream areas (secondarily affected by the policy, coefficient 0.0018) should focus on “balancing protection and development”, and strengthen quality improvement. With the characteristics of high rate, the “permanent basic farmland protection area around the city” is delimited, and illegal occupation is strictly prohibited. Thirdly, agricultural subsidy funds should be integrated, with a focus on soil improvement, water-saving irrigation, and other fields, with an annual investment of not less than RMB 750 per hectare. The grain quality rate in downstream areas should also be promoted to more than 60%. Upstream areas (weak policy effect, coefficient 0.0037) should focus on “ecology and food coordination” and explore characteristic paths. Combined with the empirical characteristics of shortage of upstream water resources and ecological fragility, first, the red line of permanent basic farmland protection should be strictly delimited, the continuous circulation of land of more than 3.33 hectares should be limited (to avoid the aggravation of the ecological pressures of large-scale operations), the model of “small land integration + special planting” should be implemented, and focus should be placed on the development of dry-resistant staple grains such as barley and crops such as potatoes. In addition to this, a policy of “ecological priority” for the withdrawal of housing estates should be implemented, where priority is given to ecological restoration of the withdrawn plots. If it really needs to be reclaimed, it should be limited to the planting of drought-resistant grain crops, and an ecological planting subsidy of RMB 22,500 per hectare should be given. Moreover, connections should be built with leading middle- and downstream agricultural enterprises and a “special grain source order planting” mechanism should be established. It is essential to ensure the income of upstream farmers from grain cultivation, ensure that the local grain self-sufficiency rate is stable at more than 90%, and ensure the stable development of food security.

6.2.2. Establish a “New-Type Urbanization + Agricultural Modernization” Synergy Mechanism to Enhance Policy Effectiveness

Based on the positive adjustment effect of new urbanization (adjustment coefficient 0.416) and agricultural modernization (adjustment coefficient 0.027), a synergistic mechanism should be established to amplify the effectiveness of reform. With the precise empowerment of new urbanization, the mismatch of land should be reduced, and the reform of household registration and housing base should be connected. For migrant workers who have paid urban social security for 3 years and have stable employment in the Yellow River Basin, the settlement restrictions should be fully liberalized. Those who voluntarily withdraw from the housing base after settling should be able to obtain the priority subscription right of urban subsidized housing, or receive per hectare at one time. A RMB 225–300,000 withdrawal compensation should be given with the goal of promoting 10% of upstream, 20% of midstream, and 30% of downstream farmers to complete the withdrawal of the housing base by 2025. The linkage between urban industries and land should be optimized, and primary processing, warehousing, and logistics facilities for agricultural products should be laid out around the city. Priority should be given to the use of collective business construction land to enter the market, the land cost of the park should be reduced, and the on-site processing conversion rate of grain should be driven to more than 50%; this research echoes the empirical conclusion new urbanization optimizes land allocation and amplifies the effect of land reform on food security. To improve the quality and efficiency of agricultural modernization, strengthen technical support, focus on the empowerment of large-scale business entities, subsidies should be given for the purchase of agricultural machinery and agricultural credit discounts to grain-growing subjects with more than 3.33 hectares of circulating land, and the comprehensive mechanization rate of arable cultivation and harvest in the middle- and downstream areas should be promoted. This should not be raised to 90%, more than 85%, and upstream for areas, to more than 70%. A special mechanism for science and technology promotion should be established, a technical team of “universities + scientific research institutes + county agricultural technology stations” should be set up, “customized” technical plans for different regions of the Yellow River Basin should be developed, and technical training should be conducted at least 4 times a year to ensure that there are 1–2 people to form a technical backbone in each administrative village; the contribution rate of agricultural science and technology progress should also be increased to more than 65%. The shortcomings of the adjustment effect of agricultural modernization should be made up for and the supporting role of reform for food security should be strengthened.

6.2.3. Establish Regional Coordination and Dynamic Monitoring Mechanisms to Ensure Policy Sustainability

In combination with the characteristics of “annual fluctuations in the policy effect” in empirical analysis, a “monitoring-evaluation-feedback” closed-loop mechanism should be built to ensure the sustainability of policies, as well as an accurate monitoring platform. In addition to this the data of natural resources, agricultural and rural areas, and statistical departments should be aggregated, and a “Monitoring Platform for Food Security and Land Reform in Counties in the Yellow River Basin” should be established. Real-time monitoring of three types of core indicators—the scale and use of land reform, changes in the quality level of arable land, grain yield and unit yield fluctuations— should be carried out, and the data should be updated every month, with monitoring reports generated every quarter. A coefficient linkage evaluation mechanism should be established, and with reference to the empirical coefficient of this study, a “reform effect evaluation index system” should be set up, and every year for each county government, the policy effect coefficient and adjustment effect coefficient should be reassessed. If the policy effect coefficient of a county is lower than the average, the policy warning should be launched and the reform measures adjusted in a targeted manner (such as increasing ecological planting subsidies in upstream areas). Protection and agricultural infrastructure construction should be settled once a year to ensure the balance of regional interests.