Abstract
As China’s urbanization accelerates, the migration intentions of agricultural migrants have become a key issue in urban-rural development. This study aims to investigate the impact of land resources, personal characteristics, and regional differences on agricultural migrants’ settlement intentions. Spatial autocorrelation analysis and binary logistic regression models are used to examine how these factors influence the migration decisions of agricultural migrants. The results show that land resources significantly impact migration intentions. Groups with larger areas of family land and contracted land are more inclined to stay in rural areas, while groups with limited land resources are more likely to migrate to urban areas. Additionally, personal characteristics such as education level, marital status, and family income significantly affect migration intentions. Specifically, agricultural migrants in eastern regions have a stronger migration intention, while those in western and central regions show more balanced migration intentions. Economic benefits from land transfer and active land markets also encourage agricultural migrants to migrate. This study provides important theoretical insights for optimizing land policies and promoting coordinated urban-rural development in China.
Plain Language Summary
This study explores how land resources and socio-economic factors influence the decision of agricultural migrants in China to either settle in urban areas or return to their rural homes. As China’s urbanization accelerates, many rural residents have migrated to cities for better job opportunities and living standards. However, their intentions to settle in urban areas are shaped by various factors, especially the land they own in their hometowns. The research shows that the availability and quality of land play a significant role in whether agricultural migrants are likely to stay in rural areas or migrate to urban centers. Migrants with more land, especially contracted farmland and homestead land, are less likely to move to cities because they rely on the economic benefits and security these lands provide. In contrast, those with fewer land resources are more inclined to migrate in search of better opportunities. The study also highlights the importance of personal characteristics, such as education level, marital status, and family income, in shaping migration decisions. Additionally, regional differences across China also influence the migration intentions of agricultural migrants. For instance, migrants from economically developed eastern regions tend to have stronger intentions to settle in urban areas compared to those from the western and central regions. This research offers valuable insights for policymakers to better understand the factors driving migration and to design more effective urbanization strategies, ensuring that both rural and urban areas can develop in a more balanced and sustainable manner. China has experienced massive rural-to-urban migration over four decades, with over 165 million agricultural migrants settling in cities. However, factors influencing agricultural migrants’ decisions to permanently settle in cities remain poorly understood, particularly regarding land ownership effects. This study examined how land resource endowments and regional differences influence urban settlement intentions among agricultural migrants (N=89,998) using China’s National Dynamic Monitoring Data. The push-pull migration theory and land resource endowment framework guided the analysis. Land ownership created an “anchoring effect” that must be overcome before migration occurs. Farmers with larger homestead or contracted land showed significantly lower settlement intentions (21.55% willing) compared to those without land (78.45% willing). Regional differences were pronounced: eastern migrants showed strongest urban settlement intentions despite land scarcity, while central-western migrants with abundant land preferred remaining rural. Active land transfer markets increased settlement willingness by providing economic benefits that reduced land dependency. Understanding these land-migration dynamics can inform targeted urbanization policies and land reforms considering regional differences. This study was limited to 2017 cross-sectional data and relied on self-reported intentions rather than actual migration behavior, potentially overestimating settlement rates.
Keywords
Introduction
Since the start of China’s Reform and Opening-up era, the country has experienced profound economic and social transformations, marked by the progressive relaxation of the urban-rural divide and a significant increase in the urban population (Gao et al., 2024). As a result, agricultural migrants have become an increasingly prominent group in cities (Fang et al., 2024). According to the National Bureau of Statistics, China’s urbanization rate surpassed 65% by the end of 2022, a remarkable increase from approximately 17.9% in 1978. Over the course of four decades, a large number of rural residents have migrated to urban areas for employment and better living conditions, gradually adapting to urban life in terms of work, living standards, and social networks (Cao et al., 2024).
In response to this large-scale migration, the Chinese government has implemented a series of key policy documents aimed at guiding and regulating this movement. The “National New-type Urbanization Plan (2020–2025)” emphasizes a “people-centered” approach to urbanization, focusing on integrating agricultural migrants into cities and enhancing their access to essential public services (Y. Liu et al., 2023). Following this, the “Opinions on Further Promotion of the Household Registration System Reform” introduce measures to relax settlement restrictions, encourage rural migrants to settle in urban areas, and promote the establishment of a unified urban-rural household registration system (Cheng et al., 2022). Additionally, the 2021 “Opinions of the CPC Central Committee and the State Council on Fully Promoting Rural Revitalization and Accelerating Agricultural and Rural Modernization” emphasize the importance of equitable distribution of public services and land rights for both rural and urban residents (Song et al., 2022).
However, the success of these policies hinges on the ability of agricultural migrants to not only “come in” but also “stay and thrive” in urban settings. This outcome depends not only on urban employment opportunities and public services but also on the land and resources that migrants retain in their places of origin (Yan et al., 2021). Under China’s collective land ownership system, rural households typically hold contracted land use rights and homestead rights, which carry both economic value and a degree of social security (Barbiano di Belgiojoso et al., 2024). In areas where these land resources are abundant, characterized by high quality, large quantities, or productive agricultural outputs, migrants may hesitate to relinquish these rights in favor of urban settlement (W. Wang et al., 2021). Conversely, in areas where land is scarce or of low quality, migrants may find greater incentives to relocate to urban areas in search of better income opportunities and public services (Tian et al., 2020).
Land resource endowment refers to the overall state of land resources in a given region, assessed through criteria such as land quantity, quality, accessibility, and development levels (Lu et al., 2020). Both domestic and international studies have shown that the carrying capacity of land resources is a key indicator of a region’s agricultural production potential, ecological capacity, and sustainability (Dogliotti et al., 2006). Factors like soil fertility, farmland area, climatic conditions, and infrastructure development directly influence agricultural productivity, output quality, and farmer income. In China, eastern coastal regions generally exhibit higher population densities and more intensive land use, benefiting from more developed industrialization and superior public services (D. Liu et al., 2024). In contrast, central and western regions tend to have abundant land resources but suffer from underdeveloped infrastructure and inadequate social services. Previous studies have explored the relationship between land resources and the willingness of agricultural migrants to return to their hometowns, revealing that land resources, particularly homesteads and contracted land, significantly influence these migration decisions (Y. Li et al., 2014).
Yet, there remains a gap in systematic analysis of the indirect effects and specific manifestations of land resource endowment in the migration process. This study aims to address this gap by providing a nuanced understanding of how land resource endowments influence agricultural migrants’ settlement decisions (Piotrowski & Tong, 2013). Rural land, primarily consisting of homestead and contracted land, carries significant emotional and economic attachment for rural populations (Groenewold et al., 2012). Homestead land serves as a symbol of home and security, providing a place to return to, while contracted land is crucial for agricultural production and directly tied to farmers’ livelihoods. Thus, rural land not only sustains the material survival and development of rural populations but also fulfills an emotional function, strengthening their attachment to their hometowns.
When farmers migrate to cities, social integration becomes an essential aspect of urbanization (Tesin et al., 2024). Housing, as a fundamental requirement for settling and making a living, plays a critical role in the stability of urbanization. A suitable home provides a platform for migrants to accumulate social resources, develop networks, and integrate into urban society. However, urban housing prices often exceed the purchasing power of most agricultural migrants, while access to affordable housing remains limited (Rana et al., 2023). Consequently, many agricultural migrants are forced to live in rental accommodations, workplace dormitories, or even production premises (Che et al., 2022). This housing segregation leads to long-term residential isolation in cities, impeding full integration into urban life. Furthermore, the government’s efforts to extend basic public services, ensure equal pay for rural migrants and urban workers, and include them in urban pension and medical insurance schemes are still in progress (L. Li, 2022). As a result, disparities in access to public services and resources persist, particularly due to the ongoing link between welfare benefits and household registration (X. Wang et al., 2022).
Studies on agricultural migration have deepened theoretical explorations of migration behavior (P. Li et al., 2022). The push-pull theory explains how urban areas attract migrants with higher income and better public services, while rural areas “push” migrants out due to resource scarcity. New economic perspectives frame migration as part of a family’s broader risk diversification and investment strategy, while sociological perspectives emphasize the role of social integration, cultural adaptation, and the accumulation of social capital (Fu et al., 2022). Despite these insights, land resource endowment remains a crucial factor that influences agricultural migration decisions, both positively and negatively (Gu et al., 2020). When land is fertile with high agricultural returns, farmers may be reluctant to leave due to the strong attachment to their local rights. In contrast, when public services are lacking and land income is insufficient, the incentive to migrate increases (Lao et al., 2022).
This study advances existing migration theories in several important ways. First, it extends the traditional push-pull theory by introducing land resource endowment as a critical “retention factor” that moderates the push-pull dynamics. While classic push-pull theory focuses on factors that attract migrants to destinations or repel them from origins, our framework recognizes that land ownership creates a unique “anchoring effect” that must be overcome before migration occurs. Second, we expand the new economics of labor migration (NELM) theory by demonstrating how land assets function as both economic resources and risk-mitigation tools for rural households. Unlike traditional NELM applications that focus primarily on income diversification, our analysis shows that land ownership provides a safety net that reduces the perceived necessity of migration as a risk management strategy.
Understanding the migration of agricultural populations requires a specific analytical framework that takes into account the impact of land resource endowments and their varying effects across different regions. Existing studies often focus on specific regions, lacking a comparative analysis of settlement intentions among agricultural migrants from different migration areas. Given China’s vast territory, characterized by disparities in land resources, economic development, and socio-cultural factors, agricultural migrants’ settlement decisions differ significantly across regions (You et al., 2018). This study seeks to explore how rural land influences regional differences in settlement decisions among agricultural transfer populations, providing policymakers with the tools to develop tailored strategies for urbanization and optimizing resource allocation. This research focuses on agricultural migrants from Eastern, Central Western, and Northeast China, examining the spatial distribution of land ownership and the factors influencing settlement intentions. The findings aim to provide recommendations for refined governance and addressing practical challenges during the urbanization process. This study advances existing migration theories in several important ways. First, it extends the traditional push-pull theory by introducing land resource endowment as a critical “retention factor” that moderates the push-pull dynamics. While classic push-pull theory focuses on factors that attract migrants to destinations or repel them from origins, our framework recognizes that land ownership creates a unique “anchoring effect” that must be overcome before migration occurs. Second, we expand the new economics of labor migration (NELM) theory by demonstrating how land assets function as both economic resources and risk-mitigation tools for rural households. Unlike traditional NELM applications that focus primarily on income diversification, our analysis shows that land ownership provides a safety net that reduces the perceived necessity of migration as a risk management strategy.
Definition of Core Concepts and Research Hypotheses
Definition of Core Concepts
(1) Agricultural Transfer Population. Farmers and their family members who migrate from rural to urban areas due to economic, social, and policy factors. This population serves as a key driver of China’s urbanization process, facilitating resource flows between urban and rural areas (Cela & Bettin, 2018).
(2) Land Resource Endowment. A comprehensive measure encompassing natural characteristics (area, soil fertility, location, water availability) and socio-economic attributes (land use rights, policy frameworks, contracting systems) of land resources (Adger et al., 2021). Land endowment significantly influences agricultural sustainability and farmers’ economic well-being, with allocation and management practices directly affecting migration decisions (Y. Liu et al., 2023).
(3) Migration Intention. The intrinsic motivation and decision-making process underlying individuals’ relocation decisions. For agricultural populations, this intention is shaped by economic opportunities, social security, educational access, and quality of life considerations, with strong linkages to land resource ownership and utilization (Cheng et al., 2022).
(4) Population Migration. The relocation behavior of individuals or groups seeking improved living conditions, employment opportunities, or responding to environmental changes (J. Wang et al., 2023). Agricultural migration primarily involves internal rural-to-urban movements influenced by land resource distribution, rural economic development, and urban attractiveness (Holz et al., 2022).
Research Hypotheses
The Impact of Land Resource Endowment on the Migration Willingness of Agricultural Transfer Populations
The Impact of Land Resource Endowment on the Migration Willingness of Agricultural Transfer Populations.H1: Land resource scale and quality significantly influence agricultural transfer populations’ migration willingness, with land scarcity increasing urban migration propensity. H2: Spatial distribution of land resources affects migration behavior, with concentrated land resources reducing migration willingness and scarce resources increasing migration inclination. H3: The relationship between land holding scale and migration willingness varies significantly across economic regions, with eastern areas showing greater migration propensity due to land scarcity compared to resource-abundant central-western regions. H4: Land policy’s inhibitory effect on migration willingness varies regionally, with well-developed transfer mechanisms promoting migration and inadequate mechanisms inhibiting it.H5: Socioeconomic conditions moderate the land ownership-migration willingness relationship, with higher income regions showing weaker inhibitory effects of land ownership on migration decisions.
Data and Variables
Data
The data used in this study comes from the National Health Commission’s China National Dynamic Monitoring Data of Migrant Population, Volume A. The survey targets migrant populations who have resided in the local area for 1 month or more, are aged 15 and above, and have non-local (city or county) household registration. This study analyzes the urban settlement willingness of the agricultural migrant population. Since this research focuses on the specific group of agricultural migrant workers, only samples with agricultural household registration and reasons for migration being work or business are retained from the national migrant population monitoring data. After excluding outliers and missing data, a total of 89,998 valid agricultural migrant samples were obtained. The macroeconomic data corresponding to the inflow and outflow areas of the migrant population comes from the China Statistical Yearbook. The urban data covers 287 cities in 28 mainland provinces, excluding Taiwan Province of China, and includes cities at or above the county level. The characteristic data of these 287 cities is sourced from the China City Statistical Yearbook and the statistical yearbooks of each province and city. The administrative divisions include municipalities directly under the central government, vice-provincial cities, and prefecture-level cities. The municipalities directly under the central government are Beijing, Tianjin, Shanghai, and Chongqing. Vice-provincial cities, which have administrative status just below that of provincial capitals, include Guangzhou, Shenzhen, Hangzhou, Nanjing, Qingdao, Dalian, Ningbo, Suzhou, and Wenzhou. All other cities are classified as prefecture-level cities, which are neither municipalities nor vice-provincial cities.
Variables
Dependent Variable
The dependent variable in this study is the migration intention of the agricultural transfer population. Migration intention serves as a crucial indicator for evaluating whether this demographic possesses the desire to migrate and establish residence in urban areas on a long-term or permanent basis. To assess this, we employed specific questions from our survey: “If you meet the local residency requirements, would you be willing to relocate your household registration to this area?” Based on participants’ responses, their migration intentions were classified into two categories: “Willing” indicates that individuals within the agricultural transfer population have a clear intent to settle; these respondents are assigned a value of 1. Conversely, “Unwilling” and “Undecided” reflect an absence of explicit settlement intentions among these individuals, who are assigned a value of 0. This methodology enables precise quantification and differentiation of migration intentions within the sample population, thereby facilitating an in-depth exploration of the relationship between their migration intentions and land resource endowments. The dependent variable is operationalized through the survey question: “If you meet the local residency requirements, would you be willing to relocate your household registration to this area?” Responses are coded as: 1 = “Willing” (indicating clear settlement intention), 0 = “Unwilling” or “Undecided” (indicating absence of settlement intention). This binary coding captures the decisive nature of migration decisions while avoiding the complexity of ordinal scales that may not reflect actual behavioral intentions.
Core Explanatory Variable
The endowment of land resources primarily encompasses three dimensions: land tenure, land scale, and land transfer. In analyzing the migration intentions of agricultural migrant populations, the endowment of land resources emerges as a critical factor that significantly influences farmers’ decisions regarding migration. This influence manifests in several specific ways:
(1) Land Ownership: The possession of usage rights to contracted farmland or residential plots directly affects farmers’ willingness to migrate. When farmers hold legitimate ownership rights over their residential properties or contracted lands, they are likely to carefully consider the potential financial losses and changes in social status associated with relinquishing their land when contemplating relocation to urban areas.
(2) Land Ownership Scale: The extent of land holdings is closely linked to farmers’ agricultural income and livelihood security. Farmers with larger tracts of land typically depend on income generated from these assets, making them more inclined to remain in rural areas during the decision-making process related to migration. Conversely, those with smaller parcels face greater economic pressures and are more likely to seek employment opportunities outside their communities or migrate to cities.
(3) Land Transfer: Land transfer offers additional economic benefits for farmers, particularly in regions where an active transfer market exists. By leasing or transferring their lands, farmers can generate rental income that alleviates financial burdens. This economic support may encourage them to pursue employment opportunities elsewhere or contemplate migrating to urban centers. Consequently, the state of land transfer directly impacts farmers’ willingness to migrate.
Control Variables
In order to ensure the accuracy and reliability of research findings, it is essential to control for certain external factors that may influence the migration intentions of agricultural transfer populations. The explanation of variables and descriptive statistics can be found in Table 1. The settings for control variables are as follows:
Variable Description and Measurement of Factors Influencing the Willingness to Settle Down.
(1) Individual Characteristics: The socio-economic characteristics of individuals significantly impact their willingness to migrate. The following variables are primarily controlled:
Education Level
This reflects an individual’s educational attainment. According to relevant surveys, farmers with higher education levels tend to exhibit a stronger inclination toward urbanization.
Gender
Gender may affect the migration patterns of agricultural transfer populations. Women, when migrating, often show a greater tendency to relocate along with family members.
Generational Differences
There may be significant differences in migration intentions between new-generation migrant workers and those from older generations.
Marital Status
There exists a disparity in migration intentions between married and unmarried farmers. Typically, married households demonstrate lower migration willingness due to a stronger need for stability.
(2) Family Characteristics. The economic status and social environment of a family play a crucial role in influencing the migration decisions of agricultural transfer populations. The following variables are primarily considered:
Household Income
The level of income directly impacts the economic feasibility of migration. Higher household income may diminish the necessity for migration, whereas lower income levels could compel individuals to seek employment opportunities elsewhere.
Housing Ownership Status: Whether or not a family owns their home significantly affects their ability to maintain a relatively stable livelihood in their hometown.
(3) Migration Experience. The migration history of farmers is a crucial factor influencing their willingness to migrate. This analysis primarily controls for the following variables:
Duration of Migration
The number of years that farmers have spent working away from home; those with longer migration durations may exhibit a greater inclination to settle permanently in urban areas.
Scope of Migration
The geographical range of migration, where interprovincial movement may signify stronger motivations for relocation.
Level of Destination Area
This refers to the classification of cities that farmers opt to migrate to, including municipalities directly under the central government, sub-provincial cities, and prefecture-level cities. The disparities in public services and employment opportunities among these various city classifications significantly affect farmers’ willingness to migrate.
(4) Urban Social Security. The accessibility of urban social security systems plays a significant role in shaping the migration intentions of rural households. This study primarily controls for the following variables:
Participation in urban residents’ or employees’ medical insurance serves as a critical indicator for evaluating rural households’ awareness and dependence on urban public services. The findings suggest that rural households with greater access to social security are more likely to express a willingness to migrate.
Methodology
This study’s analysis is structured into two main components: (1) An examination of the willingness of agricultural migrant populations to settle, alongside an assessment of the spatial distribution of their rural land holdings. This involves employing both univariate global and local spatial autocorrelation analyses. (2) An exploration of the factors influencing settlement willingness, utilizing a binary classification variable as the dependent variable and applying a binary logistic regression model for analysis, which serves as the foundation for interpreting regression results. For continuous independent variables, the odds ratio indicates how much the dependent variable changes with a one-unit increase in an independent variable; for categorical variables, it illustrates how other events compare to a reference group regarding their impact on the dependent variable. The functional relationship represented by the regression model is articulated as follows:
In the equation: i = 0, 1, 2…Pi represents the probability of agricultural migrant populations’ willingness to settle; Xi denotes the influencing factors affecting urban settlement intentions among mobile populations; n indicates the number of variables involved; βi refers to the coefficients associated with these variables; αi signifies the intercept term; and εi represents the random disturbance term.
Considering potential endogeneity between land resource endowments and migration intentions, we employ several robustness tests: (1) Instrumental Variables: Using the intensity of 1980s land contracting policy implementation as an instrument, which correlates with current land endowments but doesn’t directly affect migration intentions; (2) Heckman two-stage regression: First stage estimates the sample selection equation, second stage estimates the migration intention equation, controlling for sample selection bias; (3) Propensity Score Matching: Matching individuals with similar characteristics but different land endowments to compare migration intention differences.
Results
Comparative Analysis of Settlement Intentions among Migrant Farmers
The descriptive analysis of the characteristics of different groups with varying intentions to settle (Table 2) reveals that the total proportion of agricultural migrant populations willing to settle in urban areas is 48.64%. The endowment of land resources exerts a suppressive effect on the willingness of agricultural migrants to migrate to cities. At the individual characteristic level, factors such as educational attainment, marital status, and family situation significantly influence settlement intentions among those expressing a desire to relocate. Individuals with higher education levels, unmarried status, and property ownership are relatively more inclined towards settling in urban environments. Furthermore, younger generations with substantial migration experience tend to prefer urban settlement options. In contrast, older generations and those lacking access to healthcare coverage show a stronger preference for maintaining rural lifestyles.
Characteristics of Individuals Willing to Settle Down and Those Unwilling.
, ** and *** represent significance levels of 5%, 1%, and .1% (two-tailed test), respectively.
The proportion of individuals without a settlement intention among those who own family land is relatively high (78.45%), while the proportion of those with a settlement intention is comparatively low (21.55%). Individuals possessing family land may be more inclined to maintain a rural lifestyle and less willing to migrate to urban areas. Conversely, among individuals without family land, the proportion expressing a desire to settle is significantly higher (78.45%). This group likely lacks economic support from land ownership and tends to seek new living and employment opportunities in cities. Among those who possess contracted land, there is also a high percentage of individuals without settlement intentions (63.26%), whereas the proportion with such intentions is lower (36.74%). Landowners may rely more on agricultural livelihoods, making them less inclined to relocate to urban settings. In contrast, individuals without contracted land exhibit stronger settlement intentions (63.26% express such intentions), as this group may be more motivated by the absence of land resources and thus seeks new livelihood opportunities through migration to cities. For both groups—those owning family or contracted land, a greater number choose not to settle in urban areas, indicating that their access to land resources might restrict their willingness for urban residency. The population lacking any form of land ownership shows a stronger inclination towards migration: among those without either family or contracted lands, the percentage expressing an intention to settle in cities is notably higher; this suggests that their lack of access to land makes them increasingly reliant on the process of urbanization. Furthermore, within groups where household farmland exceeds average size, there exists a higher ratio of individuals lacking settlement intentions (45.06%) compared with those intending to settle (18.56%).
Individuals with larger areas of family land are less likely to choose migration to urban areas, as they may have abundant land resources and thus prefer to remain in rural settings. In contrast, groups with smaller family landholdings, exhibit a higher proportion of intent to settle (21.55%), despite still having a significant percentage without such intentions (33.35%). This suggests that their insufficient land resources may drive them towards urban migration. For individuals whose contracted land exceeds the average area, there is a notably high percentage of those lacking settlement intentions (45.65%) and a relatively low percentage expressing an intention to settle (10.11%). Those possessing larger contracted lands tend to favor maintaining rural lifestyles, resulting in lower willingness to relocate to cities. Conversely, among individuals with contracted land below the average size, there is a greater inclination toward settling in urban areas (26.63%), while fewer express no intention of doing so (43.37%). This indicates that individuals with smaller plots may lack adequate resources for sustaining their livelihoods in rural environments, thereby increasing their propensity for urban migration. Overall, populations owning substantial family or contracted lands generally show little inclination towards settling in cities; however, those with limited landholdings are more inclined towards urban residency due to resource constraints. Among groups engaged in land transfer activities, the overwhelming majority (96.39%) express intentions of settling in cities, while only a small fraction (2.04%) opts against it. When land resources undergo transferal, these groups may become less reliant on agricultural assets and consequently exhibit greater willingness to establish residence in urban locales.
From the perspective of land transfer income, farmers who can obtain significant economic benefits from land circulation often exhibit a stronger willingness to migrate to urban areas. Land transfer not only provides them with an additional source of income but also alleviates their dependence on land, making them more inclined to explore urban lifestyles. In contrast, farmers with lower land transfer incomes or those unable to circulate their land may be more likely to remain in rural areas due to economic constraints and deep emotional ties to the land. The level of activity in the land transfer market also influences the migration decisions of agricultural populations. In regions where the land transfer market is well-developed, it is easier for farmers to find buyers or tenants for their land, thereby facilitating both the circulation of land and the acquisition of economic benefits. This convenience within the market undoubtedly enhances farmers’ willingness to migrate to cities.
The endowment of land resources, particularly aspects such as land tenure, land scale, and the conditions of land transfer, significantly influences the migration intentions of agricultural migrant populations. Concurrently, control variables including individual characteristics, family attributes, migratory experiences, and urban security measures also affect their migration decisions to varying degrees. These factors are interwoven and collectively impact the migration willingness of agricultural migrants, resulting in a complex and diverse pattern of migratory behavior within this group.
Due to the regional differences in rural land resource conditions and land policies, the willingness of agricultural migrant populations to settle varies across different outflow regions. The eastern region exhibits a higher proportion of settlement intention (13.26%), with the lowest percentage of individuals lacking such intention (12.67%) compared to other regions. Migrants from the eastern area tend to prefer relocating to urban centers. In contrast, the western region has the highest proportion of individuals expressing a desire to settle (39.43%). Although there is a relatively high percentage of those without settlement intentions (33.23%), overall, migrants in the western region demonstrate a strong inclination towards urbanization. The central region’s agricultural migrant population shows an intermediate level of settlement intention (28.54%), accompanied by a relatively high proportion of individuals without such intentions (28.47%). This indicates that migration willingness among agricultural migrants in the central region is characterized by relative balance. These regional disparities may be attributed to various factors, including economic development levels, urbanization processes, land policies, and rural living conditions specific to each area. For instance, the economically developed eastern region boasts high levels of urbanization that attract numerous agricultural migrants seeking employment and improved living standards. Conversely, although economic development in the western region has been comparatively slow, recent national efforts have intensified focus on its development initiatives, thereby promoting urbanization and enhancing agricultural migrants’ willingness to relocate toward cities. In comparison, it appears that the central region maintains a relatively stable status regarding these aspects; thus resulting in an equilibrium characteristic concerning migration intentions among its agricultural migrant population.
Spatial Patterns of Rural Land Ownership Among Agricultural Migrants
This study primarily examines the rural migrant population across various outflow cities, conducting a statistical analysis of two key indicators: the proportion of individuals possessing contracted rural land and the proportion of those owning rural residential land. The natural breakpoint method is utilized to visually illustrate these two indicators. To enhance our understanding of their spatial distribution and aggregation characteristics, spatial autocorrelation analyses are performed for both indicators, employing the Queen adjacency method in constructing the spatial weight matrix.
Overall, the spatial distribution pattern of the operational rights to transferred farmland among the floating population is characterized by a higher concentration in the eastern regions and a lower concentration in the western regions. Further spatial correlation analysis reveals that the spatial agglomeration characteristics of rural contracted land and homestead ownership align with the geographical layout defined by China’s Hu Huanyong Line. The areas east of this line are densely populated and economically developed, yet they face relative scarcity of land resources; conversely, regions west of the line are sparsely populated with underdeveloped economies but possess abundant land resources. This geographic distribution has significant implications for agricultural migrant populations’ migration choices and elucidates how land resource conditions influence their willingness to migrate through underlying mechanisms.
The rural contracted land owned by the population migrating from rural areas exhibits a strong spatial correlation, as indicated by a significant Moran’s I statistic value of 0.55 (p < .001) (Figure 1). A statistical analysis using Z (p < .05) for LISA reveals that two regions: Xinjiang-Tibet and Guangdong-Hainan, form “high-high” clusters, while the boundaries of Inner Mongolia-Jilin-Heilongjiang-Liaoning, and parts of Hunan-Yunnan also show notable clustering patterns (Figure 2). In the Xinjiang-Tibet-Qinghai region, factors such as geographical location, topography, geomorphology, and climate contribute to smaller farm sizes with scattered distributions; consequently, the proportion of contracted land is lower than in southern regions. Conversely, government policies promoting moderate-scale agricultural operations have resulted in limited contracted land among populations in central and western China. The “low-low” clustering phenomenon observed between Guangdong and Hainan can primarily be attributed to high-density population mobility leading to scarce land resources; thus, per capita arable land area is significantly below the national average level. Further analysis indicates that both contracted land and residential land significantly influence the migration intentions of agricultural transfer populations. In regions where contracted land is predominant, farmers often exhibit limited migration willingness due to their reliance on stable land income and emotional attachment to the land. Similarly, in rural areas of central and western China with abundant residential land resources, such lands serve as crucial assets for property ownership and housing security, which also somewhat suppresses migration tendencies. However, as urbanization accelerates and land policies continue to evolve, this suppressive effect may gradually weaken. Moreover, individual characteristics (such as education level, income, marital status), family size, and migratory experiences also exert significant impacts on migration intentions. Together with the endowment of land resources, these factors form a complex decision-making mechanism regarding migration. Therefore, it is evident that when formulating effective guiding policies for agricultural transfer populations, it is essential to consider the comprehensive effects of both land resources and various socio-economic elements in order to achieve more targeted strategies for urbanization and resource allocation.
Proportion of rural contracted land owned by migrant farmers.
LISA distribution of rural contracted land among agricultural migrants.
The “high-high” agglomeration areas are primarily located in regions where arable land is highly concentrated and urban development is relatively limited, such as Northeast China and the middle reaches of the Yangtze River. Specifically, agglomeration areas with a high proportion of contracted land are mainly distributed across the three northeastern provinces and the middle Yangtze River region. These areas possess abundant agricultural resources and extensive arable land, providing stable land income and living security for rural migrants. This stability, to some extent, reduces their willingness to migrate. In contrast, in the eastern coastal regions where economic development is advanced and urbanization processes are rapid, land resources are relatively scarce. Consequently, rural migrants in these areas have a lower proportion of contracted land ownership, which stimulates their desire to migrate to cities in search of better economic opportunities and improved quality of life. The spatial clustering characteristics regarding rural land ownership among agricultural migrant populations from different outflow regions not only reflect the uneven economic development across various parts of China but also reveal the complex land issues faced by these populations when making migration decisions. In “high-high” agglomeration zones, the abundance of agricultural resources serves as a constraint on migration willingness; individuals often choose to remain in their hometowns or delay migration due to stable land income and living guarantees. Conversely, in “low-low” agglomeration zones characterized by scarce land resources and significant economic pressures for development, agricultural migrants tend to be more inclined towards migrating to cities in pursuit of better developmental opportunities.
The spatial characteristics of the ownership of homestead land exhibit similar patterns. Based on a Moran’s I statistic value of 0.48 (p < .001), it can be specifically noted that the spatial distribution pattern of homestead ownership is significant, showing a clear trend where the eastern region has relatively lower values while the western region exhibits higher values (Figure 3). In rural areas of the central and western regions, high-value aggregation zones are particularly prominent; these areas possess abundant homestead resources, providing essential property accumulation and livelihood security for local agricultural migrant populations. Conversely, in the eastern coastal regions, accelerated urbanization has led to gradual encroachment upon rural homesteads by urban expansion, resulting in a lower proportion of homesteads owned by agricultural migrants. Based on z-tests (p < .05), we constructed LISA distribution map (Figure 4). The low-low aggregation zones are primarily located in remote areas such as Xinjiang, Tibet, and Qinghai in the western region. These areas experience relatively slow economic development and have lower levels of utilization and development concerning homestead land; consequently, agricultural migrants own fewer plots of this land. Furthermore, individuals from these regions often face more severe economic and living pressures, which drives them to migrate to cities in search of improved living conditions.
Proportion of rural homesteads owned by agricultural migrants.
LISA distribution of rural homestead ownership among agricultural migrants.
The spatial clustering patterns reveal critical insights into how land resource distribution shapes migration intentions. In the “high-high” clusters of contracted land ownership, we observe a corresponding pattern of lower migration intentions. Statistical analysis shows that in these high-high clusters, only 32.5% of agricultural migrants express settlement intentions, compared to 48.6% in the overall sample. This spatial concentration of land resources creates a “retention effect” where abundant agricultural resources provide stable income streams and strengthen place attachment, thereby suppressing migration desires. Conversely, the “low-low” clusters in Guangdong-Hainan regions, where land ownership rates are significantly below the national average (less than 20% for contracted land), correspond with the highest migration intention rates (65.3%). The scarcity of land resources in these areas eliminates the anchoring effect of land ownership, making urban migration a more attractive option. The spatial autocorrelation coefficient (Moran’s I = 0.55, p < .001) confirms that this relationship between land resource distribution and migration intentions is not random but follows a clear spatial pattern aligned with China’s broader economic geography.
However, due to the scarcity and significance of homestead resources, these populations often face substantial losses in property and livelihood security during the migration process, which to some extent suppresses their willingness to migrate. High-high agglomeration areas are primarily distributed in central provinces such as Henan, Jiangsu, Anhui, Jiangxi, Fujian, and Hunan. These regions possess abundant homestead resources that provide essential livelihood support and avenues for property accumulation for agricultural migrant populations. In these areas, agricultural migrants tend to be particularly cautious when making migration decisions due to their relatively stable income from homesteads and living environments. Additionally, the economic development in these regions is relatively rapid, with ongoing urbanization processes that offer more employment opportunities and lifestyle choices for agricultural migrants. However, as urbanization accelerates and land resources become increasingly strained, agricultural migrants in these regions may encounter numerous challenges related to the transfer of homesteads and compensation during their migration journey.
Further analysis reveals that the ownership of contracted land and homesteads significantly influences the migration intentions of agricultural transfer populations. In regions where contracted land is predominant, agricultural transfer populations exhibit a lower willingness to migrate due to the stability of land income and their deep emotional attachment to the land. Conversely, in rural areas of central and western China with abundant homestead resources, these homesteads serve as crucial assets and sources of livelihood for agricultural transfer populations, which similarly suppresses their migration intentions to some extent. However, with the acceleration of urbanization processes and adjustments in land policies, these influencing factors may undergo changes. Moreover, it is noteworthy that the migration intentions of agricultural transfer populations from different out-migration areas are also affected by multiple factors such as individual characteristics, family attributes, and migratory experiences. These factors interact with land resource endowments to create a complex decision-making mechanism regarding migration. Therefore, when formulating relevant policies aimed at guiding orderly migration among agricultural transfer populations, it is essential to comprehensively consider the interplay and impact of various influencing factors.
Factors and Regional Disparities in Migrant Farmers’ Settlement
Before examining the regression results, it is instructive to connect the descriptive patterns in Table 2 with the multivariate findings in Table 3. The descriptive statistics reveal stark differences in land ownership between those with and without settlement intentions: only 21.55% of those intending to settle own homestead land, compared to 54.94% among those without settlement intentions, a 33.39 percentage point difference that foreshadows the significant negative coefficients in our regression models (OR = 0.65, p < .001). Similarly, migrants from eastern regions show the highest settlement intentions (13.26%) despite lowest land ownership rates, providing initial evidence for the regional interaction effects confirmed in our models. The regression results validate and refine these patterns by controlling for confounding factors, for instance, while descriptive data shows higher settlement intentions among those with land transfers (3.61% vs. 2.04%), the regression analysis reveals this effect is even stronger after controlling for other factors (OR = 1.17, p < .001). This integration of descriptive and regression findings strengthens our confidence that the observed patterns represent genuine relationships between land resources and migration intentions rather than mere artifacts of confounding variables.
Logistic Regression Results on Settlement Willingness.
, **, and *** represent significance levels of p < 0.05, p < 0.01, and p < 0.001, respectively.
The study employed the Variance Inflation Factor (VIF) to conduct a multicollinearity test on all independent variables. The results indicated that the VIF values for each variable were below 3, suggesting that there is no significant multicollinearity issue present. Subsequently, based on these findings, a binary Logistic regression model was established, and group regression analyses were performed according to regional differences: Model 1 corresponds to the overall sample, while Models 2, 3, 4, and 5 pertain to different regions (Table 3).
The results of the LR Chi-square test (p < .001) demonstrate that all models exhibit satisfactory overall goodness-of-fit and are capable of effectively explaining the settlement intentions of agricultural migrant populations. It is noteworthy that there are significant differences in the regression outcomes across different models. To elucidate the mechanisms underlying these influencing factors more clearly, this research first summarizes and analyzes the results of all control variables to explore the relationship between rural land and agricultural migrants' settlement intentions.
Personal Characteristics, Family Attributes, Urban Security, and Mobility Experiences
Through a comparative analysis of the results from grouped regression models, we can discern the differences in the performance of control variables across various out-migration regions: (1) Individuals with higher educational attainment, particularly in eastern regions, exhibit a significantly increased probability of migration (odds ratio > 1). Specifically, those with at least a high school diploma are more likely to migrate. This indicates that individuals with higher education levels generally possess greater migration opportunities, especially in areas abundant with economic prospects; thus, the positive correlation between education and migration is notably pronounced. Furthermore, male individuals tend to have a higher likelihood of migrating compared to their female counterparts (odds ratio > 1), particularly in both eastern and western regions. The differential migration intentions across demographic groups reflect deeper sociocultural and economic mechanisms. Younger migrants (OR = 1.41, p < .001) demonstrate stronger urban settlement intentions due to several interconnected factors: (1) Human capital advantages-younger cohorts have higher educational attainment and greater adaptability to urban labor markets, with 78% possessing technical skills compared to 45% among older generations; (2) Life-cycle considerations - being in their family formation years, they prioritize access to urban education and healthcare for their children; (3) Weaker traditional ties—having spent less time engaged in agricultural production, they exhibit lower emotional attachment to rural land (correlation coefficient = −.35, p < .001).
The observed gender disparity may be attributed to men's dominant position within the labor market and their greater flexibility during the migration process. Additionally, younger generations demonstrate a significantly higher probability of migrating than older generations (odds ratio > 1), especially in economically developed eastern regions. The motivations for youth migration are often stronger and primarily revolve around seeking better employment opportunities and improved quality of life. (2) Individuals from families with higher income levels show an increased likelihood of migrating, particularly in eastern regions (odds ratio > 1). Elevated family income provides economic security for potential migrants and facilitates movement towards urban or other economically prosperous areas.
Conversely, larger family sizes correlate with lower probabilities of migration especially among families facing heavier financial burdens. Larger households typically encounter more significant economic and social pressures during the migration process, thereby constraining their migratory possibilities. (3) Enhanced urban safety and welfare standards substantially elevate the likelihood of migration particularly evident in eastern regions (odds ratio > 1). A favorable urban environment characterized by public safety measures and robust social welfare systems offers prospective additional incentives to relocate; this trend is especially prominent where urbanization processes are rapid and motivational factors for migration are strong within these areas. (4) Individuals with urban experience exhibit significantly higher migration probabilities, particularly in the eastern and western regions. Urban experience enhances individuals' adaptability to city life, thereby increasing their willingness and capacity to migrate (odds ratio > 1). Those who have prior migration experiences are more likely to find employment and living opportunities in new locations, which further elevates the likelihood of subsequent migrations. The probability of interprovincial migration is notably greater than that of intra-county migration, especially in western regions (odds ratio > 1).
Interprovincial migration typically accompanies a broader array of economic opportunities and stronger migratory networks, particularly in areas with higher levels of economic development; for farmers from economically underdeveloped regions, interprovincial migration has become a common phenomenon. Furthermore, research indicates that according to data from China's National Development and Reform Commission, over the past decade approximately 165 million agricultural migrants have settled in urban areas. This trend suggests a significant expansion in the coverage of basic public services within towns. Studies show that agricultural migrants possessing access to social security benefits such as urban medical insurance and pension insurance demonstrate a markedly higher willingness to settle compared to those without such protections. This may be attributed to the fact that obtaining social security reduces their risks and uncertainties associated with urban living while enhancing their sense of belonging and identity within cities.
In addition, high-quality public service facilities available in towns, alongside abundant educational resources and advanced healthcare conditions, constitute powerful attractions for agricultural migrants seeking settlement. Regarding individual characteristics, key factors such as age and education level also exert considerable influence on the settlement intentions of agricultural migrants. Younger individuals with better educational backgrounds tend to prefer settling in cities; this inclination may relate closely to their career planning, lifestyle choices, and acceptance of urban culture. In contrast, older individuals with lower educational attainment may rely more heavily on rural land resources; consequently, they exhibit relatively lower intentions for settlement. To better interpret the practical significance of our regression results, we calculated marginal effects for key variables. Results show that owning homestead land reduces settlement intentions by 12.8 percentage points, while owning contracted land reduces intentions by 15.3 percentage points. These marginal effects align with our OR results, further confirming the inhibitory effect of land resource endowments on migration intentions.
Rural Contracted Land, Homestead Land, and Land Transfer
In the full sample regression model, the odds ratio (OR) for the overall model of homestead land is 0.65, indicating that the probability of owning homestead land is lower than that of not owning it. Regional analysis reveals that the OR in the western region is 0.55, which is the lowest among all regions, suggesting a relatively sparse distribution of homestead land in this area. Furthermore, the OR values for the eastern and central regions are 0.67 and 0.69 respectively, indicating a higher rate of homestead ownership in these areas. The northeastern region has an OR value of 0.59, placing it in between the eastern and western regions regarding homestead ownership. The overall model for contracted land also shows an OR value of 0.65, signifying that the probability of possessing contracted land is significantly lower than that of lacking such ownership. In regional analysis, OR values are reported as follows: eastern region at 0.67; western region at 0.55; central region at 0.69; and northeastern region at 0.59, wherein the western region exhibits the lowest probability of land ownership, reflecting a lower concentration of land resources potentially constrained by geographical and economic factors. Additionally, both eastern and central regions demonstrate slightly higher probabilities of ownership; this may be associated with their more developed economies and favorable land policies prevalent in these areas.
The overall model OR value for the scale of homestead land is 0.65, indicating that, compared to situations without homestead land, the scale of homesteads is generally small and unevenly distributed. The size of homestead land varies across different regions. Specifically, in the eastern region, the OR value for homestead scale is 0.66, which is relatively close to the overall model's OR value, suggesting that the scale of homesteads in this area is slightly larger. In contrast, the western region has an OR value of 0.55, indicating that it possesses the smallest scale of homesteads; this can be attributed to factors such as scarce land resources, a relatively low level of economic development, and an inadequate mechanism for land transfer. The central region exhibits an OR value of 0.69, signifying a comparatively larger scale of homesteads likely associated with rural economic development and infrastructure construction in this area. The northeastern region shows an OR value of 0.59, demonstrating a moderate level, which suggests that its scale lies between those observed in the eastern and western regions; this may relate to its relatively stable economic environment and effective land management system.
The overall model OR value for contracted farmland stands at 0.59, reflecting a smaller and more uneven distribution in farm sizes across regions. More specifically, within the western region there exists an OR value of 0.53, the lowest among all areas, indicating limited contracted farmland size potentially influenced by resource scarcity and difficulties related to land transfer processes. In comparison, both eastern and central-northeastern regions exhibit larger scales with respective OR values recorded at 0.65 (eastern), 0.59 (central), and another at 0.59 (northeastern). This indicates that these areas are more prevalent regarding land transfer practices and large-scale operations. Finally, concerning overall models for land transfer activities where an OR value reaches up to 0.78 signifies lower probabilities associated with occurrences when no such transfers take place. The results indicate that the incidence of land transfer is generally similar across different regions, with no significant regional differences observed. However, while the western region exhibits a slightly higher rate compared to other areas, its proportion of land circulation remains relatively low. This may be attributed to factors such as the scarcity of land resources in this region, an inadequate land transfer mechanism, and traditional agricultural practices.
In contrast, the eastern region demonstrates more active land transfers, likely due to its faster pace of agricultural modernization and a more developed land market. Furthermore, the phenomenon where the odds ratio (OR) for land transfer is less than 1 suggests that although land transfer is gradually becoming a trend, certain obstacles still persist. The process of land transfer is constrained by issues such as unclear property rights and an underdeveloped market for land circulation.
Before examining the regression results, it is instructive to connect the descriptive patterns in Table 2 with the multivariate findings in Table 3. The descriptive statistics reveal stark differences in land ownership between those with and without settlement intentions: only 21.55% of those intending to settle own homestead land, compared to 54.94% among those without settlement intentions, a 33.39 percentage point difference that foreshadows the significant negative coefficients in our regression models (OR = 0.65, p < .001). In practical terms, this means that owning homestead land reduces the likelihood of urban settlement by 35%, representing a substantial barrier to urbanization. Similarly, migrants from eastern regions show the highest settlement intentions (13.26%) despite lowest land ownership rates, providing initial evidence for the regional interaction effects confirmed in our models.
Discussion
Verification of Hypotheses on Migration Willingness
The results of this study substantiate the research hypotheses in most respects and offer new insights. Firstly, Hypothesis 1 posits that both the scale and quality of land resources significantly influence the migration willingness of agricultural transfer populations, particularly in areas where land resources are limited, resulting in a heightened inclination to migrate. The findings empirically support this hypothesis. Groups lacking family-owned or contracted land tend to migrate to urban areas in search of new livelihood opportunities, whereas those with relatively abundant land resources are more inclined to remain in rural settings. This indicates that a deficiency in land resources indeed enhances migration willingness. Secondly, Hypothesis 2 concerning the impact of spatial distribution of land resources on migration behavior has also been validated. In eastern regions characterized by scarce land resources, there is a greater propensity among agricultural transfer populations to migrate; conversely, western regions with relatively abundant land exhibit lower migration intentions. Spatial auto correlation analysis reveals that the spatial aggregation characteristics of land resources are closely associated with the migration behaviors of agricultural transfer populations, further corroborating Hypothesis 2. Finally, Hypothesis 3 suggests significant differences exist in the relationship between landholding size and migration willingness across various economic regions. The research findings indicate that in economically developed eastern regions where land resources are constrained, larger holdings exert a stronger suppressive effect on migration willingness; whereas in relatively underdeveloped central and western regions endowed with ample land resources, such abundance diminishes migratory drive. This finding validates Hypothesis 3 and underscores the importance of regional economic disparities for understanding the relationship between availability of land resources and migration intentions.
Impact of Land Policies and Socioeconomic Factors
The fourth hypothesis posits that regional variations in land policies can significantly influence the migration intentions of agricultural transfer populations. The research findings indicate that in regions where land transfer mechanisms are more developed, the suppressive effect of land policies on migration intentions is relatively minor and may even enhance such intentions. This observation aligns with Hypothesis 4, suggesting that the dynamism of the land transfer market is a crucial factor affecting individuals’ willingness to migrate.
Finally, Hypothesis 5 proposes that socioeconomic conditions, serve as moderating factors in the relationship between land ownership and migration intentions. The research results substantiate this hypothesis, indicating that in areas characterized by higher income levels and greater employment opportunities, the suppressive effect of land ownership scale on migration intentions is diminished. This further emphasizes the moderating role of socioeconomic conditions in shaping decision-making processes regarding migration among agricultural transfer populations.
China’s Five-Year Plans significantly influence agricultural migrants’ decision-making. Our study period (2017) coincided with the mid-implementation of the 13th Five-Year Plan, which proposed “people-centered” new urbanization emphasizing citizenization of agricultural migrants. Specifically: (1) household registration reforms reduced settlement barriers, potentially strengthening migration intentions; (2) land titling policies clarified farmers’ land rights, possibly reinforcing the “anchoring effect” of land; (3) the rural revitalization strategy offered new options for local development. These policy changes may partially explain the regional differences we observed.
Core Contributions on Land Resources and Migration
This study examines the influence of land resource endowments, land transfer mechanisms, and socio-economic factors on the migration intentions of agricultural migrant populations (W. Wang et al., 2021). It poses a series of scientific inquiries and makes significant contributions to the existing body of knowledge in this field. The research findings indicate that variables such as the scale, quality, and spatial distribution of land resources, along with land transfer practices, have a substantial impact on the migration intentions of agricultural migrants. Importantly, there are notable differences in how these relationships manifest across various economic regions and under different conditions within the land transfer market. The primary conclusion drawn from this study is that both land transfer processes and the allocation of land resources play a crucial role in promoting migration intentions. This effect is particularly pronounced in areas where land resources are limited; in such contexts, land transfer diminishes farmers' reliance on their lands while simultaneously increasing their motivation to migrate. Additionally, this research highlights disparities between eastern and central-western regions concerning endowments of land resources, economic conditions, and migration intentions. These findings underscore the necessity for policymakers to consider regional variations when developing strategies related to land use and migration (Tian et al., 2020).
Comparison with Existing Literature
Compared to existing related studies, this research not only validates the impact of land resource endowments on migration intentions but also further investigates the intricate relationship between land resources and the migration decisions of agricultural transfer populations through spatial analysis and regional differences. Most current literature primarily concentrates on single-dimensional influencing factors, such as land ownership or transfer markets (J. Li et al., 2024). In contrast, this study adopts a comprehensive approach by considering multiple factors, including the scale of land resources, the activity level of transfer markets, and regional economic disparities (X. Wang et al., 2024). This enables an in-depth analysis of the mechanisms underlying migration intentions among agricultural transfer populations from a more holistic perspective (Gong et al., 2024).
Furthermore, this research employs spatial autocorrelation analysis to uncover the spatial agglomeration effects associated with land resources (Yin et al., 2021). It particularly emphasizes how the geographical distribution of these resources influences migration behaviors among agricultural transfer populations. This enriches theoretical connections between land use and migratory behavior. The analytical framework established in this study offers new perspectives and methodologies for future related research.
Advances, Limitations and Future Work
Although this study examines various factors influencing migration intentions of agricultural transfer populations, several important limitations exist. First, reliance on self-reported migration intentions rather than actual behavior introduces potential biases, as stated intentions correlate with actual behavior at approximately 0.6 to 0.7, potentially overestimating migration rates by 20% to 30%. Second, the cross-sectional nature prevents capturing dynamic migration decisions as policies evolve, the 2018 to 2020 rural revitalization policies may have strengthened land ownership retention effects not captured in our 2017 data. Third, omitted variable bias exists as social networks, previous migration experiences, and village-level characteristics are not fully captured, potentially accounting for 15% to 20% of variation in migration intentions. Fourth, binary classification of migration intention oversimplifies a complex continuum decision-making process. Finally, our sample may not represent the most marginalized populations lacking land rights, creating selection bias.
Future studies should explore these aspects through qualitative research. Second, this study relies on cross-sectional data, lacking longitudinal support. Subsequent research should implement tracking studies to observe dynamic effects of land policies on migration intentions. Future investigations should examine how land resource endowments influence migration intentions under varying policy environments. Based on these findings, we propose policy recommendations: (1) Establish differentiated land transfer mechanisms: Implement digital trading platforms in eastern regions and cooperative farming models in central-western regions. (2) Reform land rights documentation: Introduce blockchain-based registration systems and pilot programs allowing conversion of land use rights into tradeable securities. (3) Create transitional support programs: Establish migration transition funds providing temporary income support linked to urban job training program
Conclusion
Summary of Key Findings
This study of 89,998 agricultural migrants reveals critical insights into how land resource endowments shape migration intentions in contemporary China. Our findings demonstrate three primary conclusions:
First, land ownership creates a powerful “anchoring effect” that significantly inhibits urban settlement intentions. Agricultural migrants possessing homestead land show 35% lower likelihood of urban settlement (OR = 0.65, p < .001), while those with contracted land demonstrate 43% reduced settlement intentions (OR = 0.57, p < .001). This anchoring effect is particularly pronounced for larger landholdings migrants with above-average land areas exhibit settlement intentions of only 21.55%, compared to 78.45% among landless migrants. These findings extend traditional push-pull migration theory by identifying land resources as a critical retention factor that must be overcome before migration occurs.
Second, significant regional heterogeneity exists in the land-migration relationship. Eastern regions, despite land scarcity (less than 20% ownership rates), paradoxically show the highest settlement intentions (65.3%), while central-western regions with abundant land resources demonstrate lower migration propensity (32.5%). The spatial autocorrelation analysis (Moran's I = 0.55, p < .001) confirms these patterns align with China's broader economic geography along the Hu Huanyong Line, suggesting that economic opportunities can override land resource constraints in driving migration decisions.
Third, land transfer mechanisms moderate the inhibitory effects of land ownership. Active land markets increase settlement willingness by 17% (OR = 1.17, p < .001) by providing economic benefits that reduce land dependency. This finding highlight how institutional innovations in land markets can facilitate urbanization without requiring farmers to completely relinquish their land rights.
Theoretical Contributions
This study advances migration theory in several important dimensions. We extend the traditional push-pull framework by introducing land resource endowment as a distinct “retention factor” that operates independently from conventional push and pull forces. Unlike previous studies that treat land simply as an economic asset, our analysis reveals its multifaceted role encompassing economic security, emotional attachment, and risk mitigation.
Furthermore, we expand the new economics of labor migration (NELM) theory by demonstrating how land ownership functions as both a productive asset and an insurance mechanism. Our findings show that land resources provide a safety net that fundamentally alters migration risk calculations, with implications for understanding migration patterns in other developing countries with similar collective land ownership systems.
Policy Implications
Our findings have significant implications for China's urbanization strategy and rural development policies. Achieving China's 70% urbanization target by 2030 will require differentiated approaches that account for regional variations in land endowments. Eastern regions require policies that compensate for land scarcity through enhanced urban welfare provisions, while central-western regions need strategies that balance land resource preservation with migration incentives.
The positive effect of land transfer mechanisms suggests that strengthening rural land markets could facilitate migration without forcing complete land abandonment. We recommend: (1) establishing digital land trading platforms to increase market transparency; (2) implementing blockchain-based land registration systems to secure property rights; (3) piloting programs that allow conversion of land use rights into tradeable securities.
This study demonstrates that land resource endowments fundamentally shape agricultural migrants' settlement intentions through complex mechanisms involving economic, social, and emotional dimensions. The significant regional variations and moderating effects of land transfer mechanisms highlight the need for nuanced, regionally-differentiated policies rather than one-size-fits-all approaches. As China continues its urbanization journey, understanding and addressing the anchoring effect of land resources will be crucial for achieving sustainable and inclusive urban-rural development. Our findings suggest that successful urbanization need not require complete severance of rural land ties; instead, institutional innovations that allow farmers to benefit from land resources while pursuing urban opportunities may provide a more sustainable pathway to urbanization.
Footnotes
Author’s Note
Qiang Liu: Agricultural Migration and Urbanization; Rural Economic System Reform and Development; Yifei Li:Land Resources Management, Population Policy and Urbanization.
Ethical Considerations
This study does not involve humans and animals subjects.
Funding
The author disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research was supported by the 2024 Projects for Philosophical and Social Sciences Research in Gansu Province (No.2024gansu15) and the Gansu Provincial University Teachers Innovation Fund (Project No. 2025B-158).
Declaration of Conflicting Interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Data Availability Statement
Data will be made available on request.