Research on the Rationality of Property Rights, Income Expectation, and the Willingness of Agricultural Transfer Population to Withdraw from Homesteads: Evidence from Hefei, China

Abstract

This study aims to explore the formation mechanism of homestead withdrawal decisions among the agricultural transfer population from the perspective of the rationality cognition of property rights. Based on 1775 valid questionnaires collected from the agricultural transfer population in Hefei, Anhui Province, China, this study constructs a binary Probit model to examine the impacts of the rationality cognition of ownership and the rationality cognition of inheritance on the willingness to withdraw from homesteads with compensation, and employs the stepwise regression method and the KHB decomposition method to test the mediation effect of income expectation on the relationship between the rationality cognition of property rights and the willingness to withdraw from homesteads with compensation. The following results indicate: (1) both the rationality cognition of ownership and the rationality cognition of inheritance exert significant positive effects on withdrawal from homesteads among the agricultural transfer population; (2) the impact of rational property rights exhibits significant heterogeneity across gender, age, share of non-agricultural income, and migration regions; (3) income expectation plays a mediating role in the relationship between the rationality cognition of property rights and the willingness to withdraw from homesteads. In conclusion, we should deepen publicity and education on homestead ownership, and build a market-oriented and diversified compensation mechanism for homestead withdrawal to enhance the rational cognition of homestead property rights and income expectation of ATP (Agricultural transfer population).

1. Introduction

Since the reform and opening up, the dual rapid development of urbanization and industrialization has attracted a large number of agricultural populations to migrate to cities and towns. Except for a decline in the total number of migrant workers in 2020 due to the COVID-19 pandemic, the national total number of migrant workers increased year by year from 2016 to 2024, reaching 299.73 million by 2024 [1]. Against this backdrop, the long-standing residential and production-supporting functions borne by rural homesteads have weakened, and the idleness and abandonment of rural homesteads have become increasingly prominent [2].

Activating and utilizing rural homesteads is a pivotal measure to raise farmers’ property income, facilitate large-scale agricultural operation and the flow of urban–rural factors, thereby underpinning people-centered new-type urbanization and rural revitalization. The Five-Year Action Plan for In-Depth Implementation of the People-Centered New-Type Urbanization Strategy issued by the State Council in 2024 explicitly sets the citizenization of the ATP(Agricultural transfer population) as the primary task of new-type urbanization [3]. Meanwhile, the Comprehensive Rural Revitalization Plan (2024–2027) issued by the Central Committee of the Communist Party of China and the State Council in 2025 stipulates that substantial progress shall be achieved in comprehensive rural revitalization and new strides shall be made in the modernization of agriculture and rural areas by 2027 [4]. Taken together, these two major strategies both aim to promote the two-way flow of urban–rural factors and optimize the allocation of urban and rural land resources. As a critical pivot linking urban and rural land resources, rural homesteads directly determine the collaborative effectiveness of the two strategies in terms of their utilization efficiency.

From a global perspective, the rapid advancement of urbanization is generally accompanied by worldwide challenges including excessive consumption of natural resources, worsening environmental pollution and escalating urban–rural social conflicts. In addition, the world is confronted with a food security crisis triggered by disruptions to agricultural production. Meanwhile, in the process of urbanization in China, there has long been a “people-land inversion” phenomenon, where the rural population declines while rural construction land does not decrease but instead increases. This has resulted in idle and wasteful rural homestead land coexisting with a severe shortage of urban land, threatening the quantitative security of arable land. In the Comprehensive Rural Revitalization Plan (2024–2027), China regards “further consolidating the foundation of national food security” as its primary objective. While pushing forward urbanization, China firmly adheres to the redline of 1.8 billion mu of cultivated land and upholds the strategic bottom line of safeguarding national food security. Accordingly, exploring effective approaches to activate and utilize rural homesteads also bears universal significance beyond national boundaries.

Institutional reform of rural homesteads in China adheres to the basic framework of the “separation of three rights”. A diversified classified disposal system targeting idle rural homesteads has been established, covering voluntary compensated withdrawal from homesteads, liberalized circulation of the right to use, and overall revitalization coordinated by village collectives. Among these approaches, the homestead withdrawal strategy can convert idle rural homesteads into reclaimed cultivated land or collectively operated construction land, which not only releases construction space for urbanization development, but also safeguards the bottom line of food security and creates favorable conditions for large-scale agricultural operation. Promoting voluntary compensated withdrawal from rural homesteads has received close attention from the Communist Party of China and the state. Policy documents including the Resolution of the Central Committee of the Communist Party of China on Further Deepening Comprehensive Reform to Advance Chinese Modernization (2024), the Comprehensive Rural Revitalization Plan (2024–2027) (2025), and the No.1 Central Document of 2026 all stress the legal protection of land rights and interests for farmers who have settled in cities [4,5,6], and call for exploring the formulation of rules for voluntary compensated withdrawal from homesteads. Accordingly, establishing a mechanism for voluntary compensated withdrawal from rural homesteads stands as the core task of rural homestead institutional reform [7,8]. Currently, abundant practical experience has been accumulated in the withdrawal from homesteads, with various localities formulating diversified pathways for such withdrawal tailored to local conditions, including the China Jinjiang Model, Yujiang Model, and Yiwu Model [9]. However, the majority of ATP still holds reservations about withdrawal from homesteads, and the advancement of rural homestead withdrawal is caught in a practical dilemma marked by strong government enthusiasm yet insufficient farmer willingness [7,9]. Accordingly, improving the willingness of ATP to WFH(Withdraw from homesteads) constitutes an urgent research topic for the in-depth implementation of the people-oriented new urbanization strategy.

Currently, there is a large body of academic research on the factors influencing the willingness to WFH. First, regarding individual and household endowment characteristics, previous studies have confirmed that factors including age, occupation, educational attainment, and income exert significant effects on the willingness to WFH [10]. Second, from the internal psychological perspective, scholars have measured the impacts of factors such as rural attachment [11], perceived value [12], risk expectation [13], and endowment effect [14] on the willingness to WFH based on large-sample survey data of rural households. Finally, in terms of the external environment, existing studies have systematically examined the degree of influence of factors including natural location [15], economic development [16], and government regulation [17] on the willingness to WFH by adopting an approach integrating theoretical analysis and empirical analysis. Furthermore, unlike most studies taking rural households as survey respondents, several studies focus on ATP as a key actor in withdrawal from homesteads [18]. Grounded in the rational decision-making perspective [19], they conduct comparative analysis focusing on the two dimensions of urban pull and rural push, and summarize the differences in willingness to WFH among ATP of different generations [20]. Overall, existing studies have explored the main factors influencing willingness to WFH from multiple perspectives and provided abundant feasible recommendations for advancing the homestead withdrawal process of farmers settled in urban areas. Nevertheless, it should be emphasized that the property rights structure determines the pattern of income distribution [21], and ATP’s cognition of rural homestead property rights directly shapes their value judgment of the costs and benefits of withdrawal from homesteads [22,23]. This form of property rights cognition among ATP exerts an ever-intensifying influence on their willingness and behavior to withdraw from homesteads [24]. However, current survey data show that many respondents’ judgments on rural homestead property rights ownership clearly contradict legal provisions [25], and farmers lack the rational cognition of property rights for homesteads [26]. This is also the core reason for ATP’s low willingness to withdraw from homesteads.

Given this, this study uses valid survey data from the 1775 agricultural transfer population in Hefei, Anhui Province, China. It employs a binary Probit model and a mediation effect test. These methods systematically investigate the effects and mechanisms of the rationality cognition of property rights on the willingness to WFH. It aims to provide references for designing incentive policies to promote homestead withdrawal among ATP. Compared with previous studies, this study may have four marginal contributions. First, this study takes ATP, a group more eligible for withdrawal from homesteads, as its survey respondents. It measures respondents’ level of rational cognition of property rights by verifying whether their understanding of rural homestead property rights aligns with the legal connotation of homestead property rights provisions. On this basis, it systematically clarifies the impact and specific mechanisms of the rationality cognition of property rights on willingness to WFH, thus enriching existing research on the driving mechanisms of such willingness. Second, this study examines the rationality cognition of ownership and the rationality cognition of inheritance. These two dimensions are prioritized most highly by ATP. The paper further analyzes their heterogeneous impacts and interaction on willingness to WFH, which serves as a valuable supplement to existing studies on homestead property rights cognition. Third, this study introduces income expectation into the analytical framework. It offers a new perspective for interpreting how property rights cognition shapes farmers’ willingness to WFH. Fourth, this study identifies group heterogeneity in the impacts of property rights rationality. Such findings improve the practical applicability of the research conclusions.

2. Theoretical Analysis and Research Hypotheses

2.1. Impact of Rationality Cognition of Property Rights on the Willingness to Withdraw from Rural Homesteads

The rationality cognition of property rights means that respondents’ cognition of rural homestead property rights is consistent with legal provisions without deviations [22]. On the one hand, Chinese law stipulates that rural homesteads are collectively owned by farmers, and village collective economic organizations exercise collective ownership on behalf of collectives. In practice, however, the principle that land follows housing is widely recognized in rural areas. Village collective economic organizations actually control only a small number of rural homesteads. Most rights of possession, use, and income of homesteads have been transferred to individual farmers [27]. This greatly restricts the space for collective economic organizations to perform management functions over homestead ownership. Farmers enjoy substantial autonomy in acquiring and occupying homesteads. This creates favorable conditions for ATP to form a privatized understanding of homestead property rights. This indicates that if policy interventions can correct such biased property rights cognition and guide it back to the rational scope stipulated by law, the WFH is expected to increase significantly. Existing studies demonstrate that stronger private ownership awareness of homesteads among farmers corresponds to lower willingness to WFH [25]. Conversely, the rationality cognition of ownership can promote such withdrawal willingness. On the other hand, the traditional inheritance perception embedded in the ancestral view of homesteads leads farmers to regard rural homesteads as a symbol of family continuity and prosperity. It endows strong ethical attributes to the intergenerational transmission of homesteads, so homestead inheritance often takes the form of division among multiple sons or is passed down within the whole clan [28]. Deep-rooted and inherited perceptions of homestead property rights held by farmers exert adverse effects on the implementation of homestead withdrawal [29].

In summary, farmers show irrational cognition regarding homestead ownership and inheritance rights. Such irrational property rights cognition reduces their willingness to WFH. By contrast, improving farmers’ rational cognition of land property rights helps advance homestead withdrawal practices. Accordingly, Hypothesis H1 is proposed:

H1. 

The rationality cognition of property rights of ATP exerts a significant positive effect on their willingness to WFH.

2.2. Moderating Role of Income Expectation

The influence of property rights cognition on income expectation has become increasingly prominent. On the one hand, the more rationally ATP understands rural homestead property rights, the more likely they are to promptly identify ultra vires acts by collective economic organizations, village committees, local governments, and other entities in the process of withdrawal from homesteads and adopt countermeasures accordingly. This lowers their perceived risk of declining land-related income [30]. On the other hand, farmers with rational cognition of property rights can gain a deeper understanding of the underlying logic behind homestead withdrawal policies. From a long-term perspective, they can recognize potential welfare and social benefits beyond monetary compensation, including supporting resources for education, recreation, and food security. This enables them to make reasonable assessments of homestead value, recognize the policy value of homestead withdrawal, and raise their income expectation [31]. However, a considerable proportion of farmers currently exhibit irrational property rights cognition [22]. Affected by the endowment effect, ATP tends to hold a privatized inclination toward homestead property rights. They regard rural homesteads in their hometown as part of their personal endowments, assign higher economic value to such land, and exhibit obvious loss aversion. Such prevalent loss aversion further raises their perceived costs of withdrawal from homesteads and reduces their income expectation regarding homestead withdrawal. To avoid potential losses triggered by homestead withdrawal, ATP will demand higher compensation corresponding to the value of their perceived land rights [32,33]. In addition, homestead compensation standards remain relatively low in most regions [34,35]. When ATP lacks rational cognition of property rights, their expected gains from participating in homestead withdrawal will decline further.

Guaranteeing reasonable income for the ATP from their withdrawal from rural homesteads constitutes an inherent requirement for advancing people-centered new-type urbanization. Income expectation is closely linked to individual behavioral decision-making. Theoretically, rural homestead withdrawal in China has both non-excludable and rival characteristics, and the resulting land appreciation gains constitute a typical closed-access common-pool resource. In current practices of withdrawal from homesteads, it is quite common to ignore key concerns and core interests of participating farmers, which undermines farmers’ principal status and sense of ownership [36]. As rational economic actors, ATP aims to maximize personal interests. They particularly worry about suffering losses if they take the initiative to withdraw first, so they tend to adopt a wait-and-see stance while hoping for higher compensation. Accordingly, they show low enthusiasm for participating in homestead withdrawal. Only when they perceive considerable expected benefits will they hold a positive attitude toward withdrawal. To resolve the collective action dilemma arising from common-pool resource governance in homestead withdrawal, it is necessary to align farmers’ short-term interest demands with the long-term national interests of land resource development. Adequate benefit compensation can offset farmers’ risk concerns and maximize the incentive effect of compensation proceeds on ATP to participate in withdrawal from homesteads [34]. For instance, Lü Junshu and Weng Xiaoyu [37] verified based on a hundred-village survey that the rationality and fairness of compensation constitute a core factor affecting farmers’ homestead withdrawal decisions. Similarly, Wang Xuebing et al. [38] drew conclusions from field surveys among farmers in China’s Shou County, confirming that income expectation exerts a significantly positive effect on willingness to WFH. In summary, continuously improving the rational cognition of homestead property rights among ATP helps raise their income expectation regarding homestead withdrawal, and thereby increases their willingness to WFH.

Accordingly, Hypothesis H2 is proposed:

H2. 

Income expectation plays a mediating role between ATP’s rationality cognition of homestead property rights and their withdrawal willingness.

H2a. 

The rationality cognition of homestead property rights exerts a significant positive effect on ATP’s income expectation of homestead withdrawal.

H2b. 

Income expectation has a significant positive impact on ATP’s willingness to WFH.

3. Materials and Methods

3.1. Data Sources and Description

The data used in this study were mainly collected through random questionnaires distributed both online and offline from October 2023 to August 2024. Survey sites covered Hefei Bus Station, Hefei South Railway Station, and Furong Community in Hefei, Anhui Province, China. Respondents needed to meet two basic criteria: they were employed in urban areas and still held homesteads in their rural hometowns. A total of 1850 questionnaires were collected, among which 1775 were valid, corresponding to an effective response rate of 95.95%. The questionnaire covered respondents’ individual characteristics, willingness to WFH, their cognition of homestead property rights, and their income expectation related to homestead withdrawal.

The descriptive statistics reveal that the gender ratio of the surveyed samples is approximately 6:4, with slightly more male respondents. Respondents aged 45 and above account for 42.03%, and the average age of the whole sample is 40 years old. Thus, the sample covers both the new generation and the older generation of ATP in a balanced manner.

In terms of educational attainment, 48.22% of respondents received higher education (college, bachelor, postgraduate), while the remaining 51.78% had no higher education (primary school and below, junior high school/technical secondary school, senior high school/higher vocational college), showing a roughly even distribution. For social insurance participation, 71.04% of respondents held three or more types of insurance, reflecting the rising penetration rate of social insurance among Chinese residents in recent years. Around 65% of households had an annual income ranging from 50,000 to 150,000 RMB, which is consistent with the monthly average income data for migrant workers released by the National Bureau of Statistics of China. In addition, roughly 60% of respondents have purchased commercial housing. Around 42% of their family members work as civil servants, teachers, or hold CPC(Communist Party of China) membership. Regarding housing types in rural hometowns, multi-story buildings take the largest proportion at 49.97%, followed by bungalows at 38.54%, and tile-roofed houses account for the smallest share of 11.49%. Overall, the sample in this study possesses certain representativeness.

Moreover, 70.14% of respondents expressed willingness to withdraw from rural homesteads. This indicates that homestead withdrawal has gained popular support among ATP to a certain extent. Policymakers should explore incentive policies to further lift withdrawal willingness and translate such willingness into actual withdrawal behaviors.

3.2. Variable Selection

3.2.1. Dependent Variable

The dependent variable refers to ATP’s willingness to WFH. It is measured by asking respondents: “Are you willing to give up your rural homestead in your hometown in exchange for monetary compensation?”

3.2.2. Core Explanatory Variable

The core explanatory variables consist of the rationality cognition of homestead ownership and the rationality cognition of homestead inheritance. A response consistent with relevant legal provisions is defined as rational cognition; otherwise, it is defined as irrational cognition.

First, the rationality cognition of homestead ownership is measured with the question: “Who do you believe owns the rural homestead in your hometown?”

Second, the rationality cognition of homestead inheritance is measured with the question: “Do you think your rural homestead in your hometown can be inherited by your children?”

3.2.3. Mediating Variable

The mediating variable is income expectation. It is measured via the survey question: “How do you evaluate the benefits you can obtain from withdrawing your rural homestead in your hometown?”

3.2.4. Control Variable

Referring to existing studies [10,11,12,13,14,15,16,17], this study sets the following items as control variables: gender, age, educational attainment, number of social insurance types participated in, annual household income, whether the household owns commercial housing, whether any family member is a civil servant, teacher, or CPC member, and the housing type of the rural homestead in the hometown.

At the stage of empirical analysis, Stata 17.0 and StataMP 18 (64-bit) are used for data collation and analysis, except that IBM SPSS Statistics 27.0 is adopted for the common method bias test. The descriptive statistics of the variables in this study are presented in

Table 1. Variable definitions and descriptive statistics.

Variable Type	Variable Name	Definition and Assignment	Mean	Standard Deviation
Dependent Variable	willingness to withdraw from rural homesteads	0 = unwilling; 1 = willing	0.701	0.458
Core Independent Variable	The rationality cognition of ownership	0 = irrational cognition; 1 = rational cognition (Belonging to rural collective)	0.609	0.488
	The rationality cognition of inheritance	0 = irrational cognition; 1 = rational cognition (Cannot inherit the homestead)	0.145	0.353
Mediating Variable	income expectation	1 = very low benefit; 2 = relatively low benefit; 3 = moderate benefit; 4 = relatively high benefit; 5 = very high benefit	3.462	1.097
Control Variable	Gender	0 = Female; 1 = Male	0.608	0.488
	Age	1 = 18 to 24 years; 2 = 25 to 34 years; 3 = 35 to 44 years; 4 = 45 to 54 years; 5 = 55 years and above	3.074	1.103
	Educational attainment	1 = primary school and below; 2 = junior high school/technical secondary school; 3 = senior high school/higher vocational college; 4 = college/bachelor; 5 = postgraduate	3.185	1.018
	Number of participants in social insurance	ranging from 0 to 5	3.435	1.525
	Annual Household Income	1 = 50,000 CNY (Chinese Yuan) or fewer; 2 = (50,000, 100,000] CNY; 3 = (100,000, 150,000] CNY; 4 = (150,000, 200,000] CNY; 5 = more than 200,000 CNY	2.592	1.113
	Whether to Buy Commercial Housing	0 = No; 1 = Yes	0.599	0.490
	Whether any Household Member is a Public Servant, Teacher or CPC Member	0 = No; 1 = Yes	0.419	0.493
	Type of Housing in Rural Hometown	1 = tiled-roofed houses; 2 = bungalows; 3 = multi-story buildings	2.385	0.683

Note: In 2024, the average annual exchange rate is 1 CNY = 0.140 USD (United States Dollar).

.

3.3. Model Specification

The dependent variable in this study is a binary choice variable. Accordingly, this paper adopts a binary Probit model to estimate the effect of property rights rationality on homestead withdrawal willingness. The baseline model is specified as follows:

$$prob(Willingness=1|x)=\varphi ({\alpha }_1+{\beta }_{1}OR+{\beta }_{2}IR+{\beta }_{3}Control)$$

(1)

In Equation (1), $Willingness$ denotes the willingness to WFH; ${\alpha }_1$ is the constant term; ${\beta }_1$, ${\beta }_2$, ${\beta }_3$ are coefficients to be estimated; $OR$ represents ownership rationality, $IR$ represents inheritance rationality, and $Control$ stands for the set of control variables; $\varphi (·)$ refers to the cumulative distribution function.

Subsequently, the stepwise regression test method is adopted to construct the following three equations.

$$IE={\alpha }_2+{\beta }_{4}OR+{\beta }_{5}Control+{\epsilon }_1$$

(2)

$$IE={\alpha }_3+{\beta }_{6}IR+{\beta }_{7}Control+{\epsilon }_2$$

(3)

$$Willingness={\alpha }_4+{\beta }_{8}OR+{\beta }_{9}IR+{\beta }_{10}IE+{\beta }_{11}Control+{\epsilon }_3$$

(4)

In Equations (2)–(4), ${\alpha }_2$, ${\alpha }_3$, ${\alpha }_4$ are constant terms, ${\beta }_4$–${\beta }_{11}$ are coefficients to be estimated, ${\epsilon }_1$,${\epsilon }_2$,${\epsilon }_3$ represent error terms, and $IE$ denotes the mediating variable income expectation. Equation (2) reflects the impact of ownership rationality on the mediating variable income expectation; Equation (3) examines how inheritance rationality affects income expectation; Equation (4) estimates the direct effects of ownership rationality, inheritance rationality, and income expectation on homestead withdrawal willingness.

In addition, the KHB model can address scaling issues arising from different variable types between mediators and dependent variables [39]. Therefore, this study applies the KHB model to decompose the magnitude of the mediating effect.

3.4. Common Method Bias Test

Harman’s single-factor test was adopted to test for common method bias, incorporating the explained variable, core explanatory variables and mediating variable into the analysis. The results identified two factors with eigenvalues greater than 1. The maximum factor accounted for 33.967% of the total variance, which was lower than the critical threshold of 40%. Accordingly, serious common method bias does not exist in this study.

4. Empirical Findings and Analysis

4.1. Benchmark Regression

Model estimation results are presented in

Table 2. Regression results of the influence of property right rationality on homestead exit intention.

Variable Name	Coefficients	Standard Errors	Marginal Effect
Ownership Rationality	0.176 ***	0.066	0.058 ***
Inheritance Rationality	0.315 ***	0.098	0.105 ***
Gender	−0.202 ***	0.066	−0.067 ***
Age	0.031	0.032	0.010
Education Attainment	−0.016	0.038	−0.005
Number of Participants in Social Insurance	0.113 ***	0.023	0.038 ***
Annual Household Income	0.022	0.032	0.007
Whether to Buy Commercial Housing	0.114 *	0.070	0.038 *
Whether any household member is a public servant, teacher or CPC member	0.188 ***	0.069	0.063 ***
Type of Housing in Rural Hometown	−0.053	0.048	−0.018
_cons	0.014	0.215
Model Observed Indicators	Log likelihood = −1040.876
	Prob > chi2 = 0.0000

Note: *** and * indicate significance at the 1% and 10% statistical levels, respectively.

. First, ownership rationality exerts a significantly positive effect on homestead withdrawal willingness at the 1% statistical level. A one-unit increase in ownership rationality raises the probability that ATP chooses to WFH by 5.8%. Second, inheritance rationality also significantly and positively promotes homestead withdrawal at the 1% level. A one-unit rise in inheritance rationality increases the withdrawal probability by 10.5%, indicating that inheritance rationality imposes a stronger influence on withdrawal willingness compared with ownership rationality. Collectively, Hypothesis H1 is verified.

In addition, several control variables, including gender, number of social insurance types, home purchase status of commercial housing, and family membership of civil servants, teachers, or CPC members, show significant impacts on ATP’s homestead withdrawal willingness. Specifically, female respondents, those with more insurance policies, households having civil servants, teachers, or CPC members, and families owning commercial housing tend to have a stronger willingness to WFH.

4.2. Robustness Tests

This study conducts robustness tests through two approaches: adding control variables and model replacement. (1) Adding control variables: the perceived opportunity for job promotion and career development is incorporated into the set of control variables, and the binary Probit regression is re-estimated accordingly. (2) Model replacement: The regression is rerun using a binary Logit model. As shown in

Table 3. Robustness test results.

Variable Name	Adding Control Variables		Model Replacement
	Coefficients	Standard Errors	Coefficients	Standard Errors
Ownership Rationality	0.172 ***	0.066	0.287 ***	0.109
Inheritance Rationality	0.317 ***	0.098	0.539 ***	0.170
Control Variables	Controlled		Controlled
Prob > chi2	0.000		0.000
Log likelihood	−1035.351		−1040.862

Note: *** indicates significance at the 1% statistical level.

, after substituting the core explanatory variable and switching the regression model, both ownership rationality and inheritance rationality still exert statistically significant effects on homestead withdrawal willingness with consistent coefficient signs. This confirms the reliability of the baseline regression results in this paper.

4.3. Endogenous Processing

Referring to the studies by Jia Yajuan et al. [40] and Gai Hao et al. [41], this paper constructs a new variable denoted as property rights rationality by calculating the average value of ownership rationality and inheritance rationality, and selects “whether the respondent has experienced land expropriation” as the instrumental variable.

The core of land expropriation consists of the transfer of land ownership. ATP affected by expropriation will inevitably form certain property right cognition to varying degrees in this process. Land expropriation is inherently closely associated with the rationality cognition of property rights of the ATP and can produce extended effects by shaping their property right cognition. Furthermore, “whether having experienced land expropriation” is an objective characteristic of the ATP. Existing studies have demonstrated that farmland expropriation experience exerts no influence on farmers’ actual withdrawal behavior of rural homesteads [42]. As willingness to withdraw from homesteads acts as the primary driving force for the occurrence of homestead withdrawal behavior [31], scholars adopting the Theory of Planned Behavior have found that the determinants of homestead withdrawal willingness and actual withdrawal behavior are generally consistent [43], which to some extent verifies that land expropriation experience has no direct correlation with willingness to withdraw from homesteads. In conclusion, the indicator of land expropriation experience meets the relevance and exogeneity criteria required for a valid instrumental variable.

Accordingly, this study adopts an IV-Probit model for estimation. As reported in

Table 4. Endogenous treatment results.

Variable Name	First-Stage: RCPR (The Rationality Cognition of Property Rights)	Second-Stage: WFH
	Coefficients	Coefficients
Land Expropriation	0.042 ***
RCPR		4.540 ***
Control Variables	Controlled	Controlled
N	1775	1775
AR	12.68 ***
Wald	6.60 ***

Note: *** indicates significance at the 1% statistical level.

, the p-values of the AR test and Wald test are both statistically significant, ruling out the problem of weak instrumental variables. The first-stage regression results reveal that land expropriation experience exerts a significant impact on property rights rationality. In the second-stage regression, after addressing potential endogeneity bias, property rights rationality still presents a significant effect on homestead withdrawal willingness, further verifying the robustness and reliability of the baseline regression results.

4.4. Heterogeneity Analysis

This study further examines whether the impacts of property rights rationality on homestead withdrawal willingness vary across subgroups of ATP grouped by gender, age, share of non-agricultural income, and migration scope, with results displayed in

Table 5. Heterogeneity analysis results.

Variable Name	Gender		Age		Rural Household Differentiation		Migration Heterogeneity
	Female	Male	Under 45	45 or Above	Low Share of Non-Agricultural Income	A Very High Share of Non-Agricultural Income (Above 70%)	Intra-Provincial Migration	Inter-Provincial Migration
Ownership Rationality	0.294 *** (0.107)	0.101(0.083)	0.263 *** (0.085)	0.034 (0.104)	0.134 (0.097)	0.172 * (0.091)	0.205 *** (0.077)	0.085 (0.127)
Inheritance Rationality	0.286 * (0.160)	0.341 *** (0.125)	0.474 *** (0.140)	0.169 (0.141)	0.177 (0.134)	0.454 *** (0.146)	0.231 ** (0.116)	0.550 *** (0.192)
Control Variables	Controlled	Controlled	Controlled	Controlled	Controlled	Controlled	Controlled	Controlled
N	695	1080	1029	746	884	891	1278	497
LR chi2	38.95	44.53	52.27	46.56	50.10	49.45	67.94	27.66
Prob > chi2	0.000	0.000	0.000	0.000	0.000	0.000	0.000	0.002

Note: The values in parentheses are standard errors; ***, **, and * indicate significance at the 1%, 5%, and 10% statistical levels, respectively.

. First, in terms of gender, both ownership rationality and inheritance rationality exert significant positive effects on female migrants’ willingness to WFH, whereas males’ withdrawal willingness is only significantly affected by inheritance rationality. Second, regarding age, ownership rationality and inheritance rationality are both statistically significant for respondents younger than 45 years old, while no significant effects exist among those aged 45 and above. Third, from the perspective of rural household differentiation, the two dimensions of property rights rationality only generate significant positive influences on the withdrawal willingness of migrants whose non-agricultural income accounts for more than 70% of total household income. Finally, in terms of migration heterogeneity, ownership rationality produces a larger effect size for migrants moving within the same province, while inheritance rationality shows a more pronounced and significant impact on those migrating across provinces.

4.5. Interaction Effect Test

This study further analyzes the interactive effect between homestead ownership rationality and inheritance rationality on homestead withdrawal willingness. As shown in

Table 6. Interaction between ownership rationality and inheritance rationality.

Variable Name	Coefficients	Standard Errors
Ownership Rationality	0.204 **	0.066
Inheritance Rationality	0.237 **	0.103
Ownership Rationality $\times$ Inheritance Rationality	0.547 **	0.226
Control Variables	Controlled
Prob > chi2	0.000
Log likelihood	−1037.960

Note: ** indicates significance at the 5% statistical level.

, after introducing the interaction term of ownership rationality and inheritance rationality, ownership rationality, inheritance rationality, as well as their interaction term all exert significant positive impacts on homestead withdrawal willingness. It indicates that inheritance rationality can amplify the promoting effect of ownership rationality on homestead withdrawal willingness. After introducing the interaction term, the statistical significance of ownership rationality and inheritance rationality on the willingness of homestead withdrawal decreased compared with the baseline regression without interaction terms. The underlying reason lies in the real policy context: the formation of agricultural transfer population’s homestead withdrawal willingness is jointly driven by the intertwined effects of ownership rationality and inheritance rationality. The baseline model without interaction terms merely treats the two dimensions as mutually independent variables and fails to capture their inherent correlation and interactive influence. Moreover, the introduction of interaction terms leads to mild multicollinearity, which raises the standard error of main effect coefficients and further results in a decline in the significance level.

4.6. Mediating Effect Test

The results in

Table 7. Regression results of property rights rationality, income expectation and homestead withdrawal intention.

Variable Name	Model 7-1 (Income Expectation)	Model 7-2 (Income Expectation)	Model 7-3 (WFH)
Ownership Rationality	0.231 ***		0.133 **
Inheritance Rationality		0.307 ***	0.258 ***
Income Expectation			0.191 ***
Control Variables	Controlled	Controlled	Controlled

Note: *** and ** indicate significance at the 1% and 5% statistical levels, respectively.

show that both ownership rationality and inheritance rationality have significant positive effects on income expectation, which verifies Hypothesis H2a. Meanwhile, ownership rationality, inheritance rationality, and income expectation all exert significant direct effects on homestead withdrawal willingness, supporting Hypothesis H2b. It can be concluded that income expectation plays a mediating role between property rights rationality and homestead withdrawal willingness.

In addition, the KHB method is adopted to decompose the magnitude of effects along the mediating paths. As presented in

Table 8. Test results of intermediary effect based on KHB method.

Path Classification	Effect Classification	Effect Size	p-Value
Path 1: Ownership Rationality → Income Expectation → Homestead Withdrawal Willingness	Total Effect	0.206	0.002
	Direct Effect	0.160	0.015
	Mediation Effect	0.046	0.000
Path 2: Inheritance Rationality → Income Expectation → Homestead Withdrawal Willingness	Total Effect	0.343	0.000
	Direct Effect	0.286	0.003
	Mediation Effect	0.057	0.001

, the total effects, direct effects, and mediating effects of Path 1 and Path 2 are all statistically significant. Income expectation exerts a significant partial mediating effect between property rights rationality and homestead withdrawal willingness, which confirms Hypotheses H1 and H2. In terms of effect magnitude, the direct effect is larger than the mediating effect.

5. Discussion

Promoting the withdrawal of rural homesteads by ATP is a crucial measure to revitalize idle land, raise farmers’ property income, facilitate large-scale agricultural operation and the flow of urban–rural factors, and thereby advance new-type urbanization and rural revitalization. This study incorporates property rights rationality and income expectation into the analytical framework and systematically explores the formation mechanism of homestead withdrawal decisions among ATP from the perspective of property rights cognition. The research findings are as follows:

(1)

Both ownership rationality and inheritance rationality exert significantly positive effects on ATP’s homestead withdrawal willingness, with a larger effect magnitude for inheritance rationality. The reasons may be as follows. In terms of ownership rationality, once migrants recognize that rural homesteads are collectively owned rather than privately possessed, their sentimental attachment to private homestead ownership declines. Given the government’s commitment to the principle of voluntary and compensated homestead withdrawal, those with rational cognition of homestead ownership tend to monetize such collectively owned land to facilitate their urban integration. Accordingly, they are more inclined to accept and cooperate with homestead withdrawal policies, achieving a win-win outcome for individual interests and social benefits. As for inheritance rationality, deeply rooted traditional culture makes farmers attach great importance to intergenerational land inheritance. When ATPs realize that homestead land itself is not inheritable, and village collectives have the right to recover homestead land if the attached houses collapse or disappear, their perception of homesteads as assets sustaining family continuity will be weakened. Aware that homesteads cannot be passed down permanently across generations, migrants are more willing to convert homestead rights into monetary compensation in a timely manner. They may invest such property income in children’s education, personal entrepreneurship, urban housing purchase, and other profitable activities, realizing intergenerational resource reallocation that shifts their mindset from land preservation to value investment. Meanwhile, ownership rationality and inheritance rationality interact to affect homestead withdrawal willingness. A potential explanation is that when migrants hold rational cognition of both ownership and inheritance rights, they can more clearly recognize that homestead land is not private property and cannot be inherited by their children. Accordingly, they become more willing to participate in homestead withdrawal and realize intergenerational inheritance by discounting the economic value of homesteads. The above findings are consistent with existing studies confirming that rational property rights cognition helps correct farmers’ biased perceptions of land ownership and improves their policy responsiveness to homestead system reform [44,45]. In addition, variables including gender, number of social insurance types, commercial housing ownership, and household membership of civil servants, teachers, or CPC members exert significant influences on ATP’s homestead withdrawal willingness. In terms of gender, male migrants are less inclined to WFH. A plausible explanation lies in the prevalent patrilineal inheritance tradition in rural areas: males regard retaining homesteads and rural houses as a family obligation and an indispensable foundation for clan continuity, thus showing a weaker willingness to give up their homesteads. Regarding insurance participation, more types of social insurance correspond to a stronger willingness to withdraw from homesteads. With broader social insurance coverage, agricultural transfer populations rely less on the social security function of homesteads and worry less about potential risks caused by homestead withdrawal, making them more willing to exit their homesteads. Households with civil servants, teachers, or CPC members demonstrate higher withdrawal willingness. Such individuals generally have a better understanding of national macro policies. Guided by professional ethics, they tend to play an exemplary role and respond more actively to national policy initiatives on WFH. Moreover, migrants who have purchased commercial urban housing show a stronger willingness to exit homesteads. Home purchase serves as an indicator of a higher degree of citizenization, implying their reduced dependence on the residential function of rural homesteads. Furthermore, mortgage pressure associated with commercial housing purchases motivates them to obtain monetary compensation via homestead withdrawal to ease financial burdens.

(2)

Mechanism analysis reveals that both ownership rationality and inheritance rationality exert significant positive effects on income expectation. A possible reason is that ATP with rational property rights cognition is more likely to abandon the mindset that homesteads are private assets inheritable across generations for the whole family, accompanied by a weaker perception of loss aversion and input costs when facing homestead withdrawal. Meanwhile, such property rights rationality helps them gain an upper hand in multi-stakeholder interest games during homestead withdrawal, reasonably safeguard their legitimate rights and interests, and thereby raise their income expectation. Furthermore, ownership rationality, inheritance rationality, and income expectation all present statistically significant direct effects on homestead withdrawal willingness. The significantly positive impact of income expectation can be explained by the assumption that ATP are rational economic agents: they become more willing to WFH if they anticipate considerable gains from withdrawal. This finding supplements existing literature by verifying that homestead withdrawal decisions are also constrained by individuals’ judgment of future earnings. Previous studies have mainly explored farmers’ homestead-related decisions and behaviors from perspectives such as institutional cognition, property rights arrangements, and compensation mechanisms [46,47]. They confirm that clear property rights cognition and sound institutional expectations strengthen farmers’ perception of property rights, alleviate worries over transaction risks in withdrawal, and further impose significant influences on land transfer and disposal decisions [48,49]. Nevertheless, existing research pays insufficient attention to the psychological cognitive mechanism centered on subjective income judgment in ATP’s homestead withdrawal choices.

(3)

Heterogeneity analysis indicates that ownership rationality generates stronger effects among female migrants, migrants under 45 years old, and those migrating within the province. By contrast, inheritance rationality exerts a more pronounced influence on male migrants, migrants under 45 years old, migrants with a high proportion of non-agricultural income, and inter-provincial migrants. The underlying reasons are elaborated as follows. First, with regard to gender, in traditional rural society, the inheritance of family properties such as land and houses mainly relies on intergenerational succession between fathers and sons [50], while women are excluded from ancestral property inheritance. It is widely taken for granted that sons inherit family assets from their fathers. If descendants fail to preserve inherited family properties, they may be labeled as unfilial or prodigal [51]. Consequently, women may have passively accepted the notion that homesteads do not belong to them personally, whereas male ATP tend to attach greater importance to inheritance rights when deciding on homestead withdrawal. Second, in terms of age, older cohorts of ATP generally have weaker awareness of property rights due to limited educational attainment in their youth. Furthermore, deeply ingrained traditional notions, including paternal inheritance and the desire to return to one’s hometown in old age, create strong emotional attachments to rural land, making their homestead withdrawal decisions less responsive to property rights rationality. Third, in terms of rural household differentiation, ATP with a high share of non-agricultural income usually feature a higher level of citizenization and rely less on the residential security function of rural homesteads. Once they form rational cognition of homestead ownership and inheritance rights, they are more inclined to WFH for compensation to improve their quality of urban life. Fourth, regarding migration heterogeneity, migrants moving across provinces maintain relatively weak connections with their rural hometowns and are less sensitive to homestead ownership, so their decisions are mainly driven by inheritance rationality. Due to geographical distance, cross-provincial migrants face relatively high costs to exercise and safeguard inheritance rights over houses attached to homesteads. Influenced by the concept of ancestral property, some cross-provincial migrants regard homestead inheritance as a family obligation and choose to retain their rural homesteads. In contrast, those holding rational perceptions of homestead inheritance rights tend to opt for homestead withdrawal after weighing relevant costs. Therefore, this study not only deepens the theoretical understanding of the internal formation mechanism of agricultural transfer population’s willingness to exit homesteads, but also clarifies the heterogeneous impacts of ownership rationality and inheritance rationality on agricultural transfer populations with different characteristics. It provides micro-level empirical evidence for the refined formulation and targeted implementation of homestead withdrawal policies.

(4)

It should be clarified that compensated withdrawal from rural homesteads is not a universal solution to address the inefficient utilization of rural land. Instead, it targets the ATP who have completed urban migration and no longer rely on the residential security function of rural homesteads, serving as an important component of the classified revitalization system under China’s institutional framework of the “separation of three rights” for rural homesteads. Nevertheless, not all rural residents are inclined to withdraw from rural homesteads. Targeted guidance should be provided for three categories of groups to encourage them to settle permanently in rural areas and maximize the productive efficiency of land resources. The first group comprises full-time agricultural operators represented by large-scale breeding households, family farm operators and specialized farmers. Their legitimate rights and interests pertaining to rural homesteads and contracted farmland ought to be protected to support them in developing appropriately scaled agricultural operations. The second group refers to returning entrepreneurs including migrant workers who have returned home, college graduates, and demobilized military personnel, who can leverage their own rural homesteads and rural housing to develop rural cultural tourism, agricultural product processing, rural e-commerce and other business forms. The third group consists of rural public service and governance practitioners such as village cadres, rural teachers, village doctors and grassroots agricultural technical staff, who can be supported to stay and work in villages through property right protection and fiscal subsidies. In addition, achieving a win-win outcome for both new-type urbanization and the efficiency improvement in rural land requires efforts in the following three dimensions: (1) Strictly uphold the redline of cultivated land protection. Construction land reclaimed from withdrawn rural homesteads shall be prioritized for restoration into cultivated land. Comprehensive territorial land consolidation shall be adopted to optimize the contiguous layout of farmland, create favorable conditions for large-scale and mechanized agricultural operation, and offset the pressure of cultivated land occupation amid urbanization. (2) Pursue intensive and connotative urbanization. Urban development boundaries shall be rigidly delineated to curb disorderly sprawl of cities, and shift the orientation of urbanization from land-centered expansion to people-centered urbanization. (3) Smooth the two-way circulation of urban and rural factors. On the premise of land use regulation, promote the market entry of collectively operated construction land, and feed the increment in land value back into agriculture and rural areas. Meanwhile, guide urban capital and technologies to flow into the countryside to empower rural industries and activate the endogenous value of rural land.

It should be noted that this study has certain limitations. (1) The empirical data are mainly derived from valid questionnaires collected from ATP passing through Hefei, Anhui Province, China. This research fails to conduct a comparative analysis across eastern, central, and western regions of China. The relatively small sample size and insufficient regional coverage weaken the representativeness of the sample to a certain extent. (2) The formation and evolution of homestead withdrawal willingness are inherently complex. There may exist other potential influencing factors, while the variables and corresponding indicators selected in this paper cannot fully cover all determinants of homestead withdrawal willingness. (3) This study only adopts structured questionnaires for quantitative research. The explained variable is set as a binary multiple-choice item, and open-ended written explanations and qualitative interview materials regarding respondents’ choices on homestead withdrawal willingness are not collected simultaneously. Hence, many implicit factors that are difficult to quantify, such as local attachment, family concepts, local customs, and personal risk preferences behind farmers’decisions, cannot be fully clarified. (4) Some heterogeneous results in the heterogeneity analysis may result from random coincidence, which can be further verified by in-depth research.

6. Conclusions and Policy Implications

Based on valid questionnaire data of 1775 agricultural transfer population samples in Hefei, this paper constructs a binary Probit model to test the impacts of ownership rationality and inheritance rationality on homestead withdrawal willingness. The stepwise regression method and the KHB model are further adopted to examine the mediating effect of income expectation between property rights rationality and homestead withdrawal willingness. The empirical results are summarized as follows: (1) Both homestead ownership rationality and inheritance rationality exert significant positive effects on ATP’s homestead withdrawal willingness, with heterogeneous effects across different groups; the influence magnitude of inheritance rationality is larger. Meanwhile, ownership rationality and inheritance rationality interact to affect homestead withdrawal willingness. (2) Income expectation plays a statistically significant mediating role between homestead property rights rationality and homestead withdrawal willingness. (3) Control variables, including gender, number of social insurance types participated in, and commercial housing purchase status, also impose significant influences on ATP’s willingness to WFH.

Based on the above research conclusions, the following policy recommendations are proposed:

(1)

Deepen public education on homestead property rights cognition. On the one hand, the government should adopt mobile legal popularization to bridge the “last mile” of legal publicity. By regularly providing door-to-door legal services and organizing touring lectures on homestead property rights, ATP can gain an in-depth understanding of the statutory attributes of rural homesteads, namely that “possession does not equal ownership” and that homestead land itself is not inheritable. Such initiatives help foster rational cognition of homestead property rights among ATP. Meanwhile, local authorities should publicize legal remedies for homestead property rights infringement. This enables ATP to rationally apply professional property rights knowledge to safeguard their legitimate rights and interests during homestead withdrawal, and reduces their perceived benefit and risk concerns regarding homestead withdrawal behavior. On the other hand, targeted education should be carried out to empower key groups. For instance, since the property rights rationality exerts a stronger influence on the new generation of ATP, who will gradually gain greater decision-making power over homestead withdrawal, basic knowledge regarding rural land policies and citizens’ property rights can be incorporated into the curricula of schools with a large number of rural-origin students. In addition, publicity approaches can be enriched in line with young people’s information acquisition habits. Digital communication modes including interactive short videos, prize-based mini-program quizzes and special live-streaming sessions can be utilized to boost the new generation’s enthusiasm for learning knowledge related to homestead ownership.

(2)

Establish a market-oriented and diversified compensation mechanism for homestead withdrawal. On the one hand, it is necessary to improve the pricing evaluation system for rural homesteads and promote the construction of market trading platforms for rural homestead property rights. First, professional rural real estate appraisal teams can be hired to conduct a homestead valuation. Government-public interactive platforms should be built to encourage participation by ATP, so as to scientifically formulate the benchmark land price system for rural homesteads. Second, on the basis of improving offline service stations for rural property rights transfer, an integrated online transfer platform covering cities and counties should be developed to realize full-process online administration and transaction operations. On the other hand, diversified compensation schemes for homestead withdrawal should be explored and implemented by fully considering the differentiated interest demands of ATP. First, investigate and identify heterogeneous demands for homestead withdrawal among groups varying in age bracket, employment type, and migration range. Second, roll out multiple compensation modes, including monetary compensation, in-kind compensation, and rights-and-interests compensation, to satisfy farmers’ diverse demands, such as cashing out benefits in the short term, securing long-term social guarantees, and reserving the option to return to their hometowns, thus avoiding social risks arising from a one-size-fits-all policy. Third, standardize compensation procedures for homestead withdrawal, boost transparency of relevant compensation information, and introduce farmer supervision and feedback mechanisms to guarantee fairness throughout the whole compensation process.