Does Labor Transfer Improve Farmers’ Willingness to Withdraw from Farming?—A Bivariate Probit Modeling Approach

Abstract

Because of the increased expansion of the non-agricultural industry spurred on by vigorous urbanization, labor migration or transfer from farm to urban regions is to become more predominant in China. Studying the effect of labor transfer on farmers’ willingness to withdraw from land is conducive to deepening the understanding of the reality of the “separation of human and farmland”. As most rural livelihoods, directly and indirectly, depend upon farming, the socio-economic impact of leaving the homestead fosters profound research value. Moreover, it would provide a decision-making reference for the government to improve the design of the rural land withdrawal system and related support policies. This article uses the survey data of 953 farmers in Shaanxi, Sichuan, and Anhui, China, to empirically analyze labor transfer’s effect on farmers’ willingness to withdraw from farmland. We construct a bivariate Probit model by eliminating the endogenous issue to craft its findings. This study outlines its findings: (i) 61.805% of the farmers were unwilling, and 18.048% were willing to withdraw from the contracted land and homestead. While 12.067% of the farmers were only willing to withdraw from the contracted land, 8.080% of the farmers were only willing to withdraw from the homestead. Further testing found a positive correlation between farmers’ willingness to withdraw from contracted land and the homestead. (ii) The overall labor transfer of households can increase the willingness of farmers to quit contracted land and homestead farming. The incomplete labor transfer of households can improve the willingness of farmers to quit contracted land. Still, it has no significant impact on the willingness of farmers to quit their homesteads. The family’s complete labor transfer incentivizes farmers’ willingness to withdraw from contracted land and the homestead, which is more potent than incomplete family labor transfer. (iii) Incomplete labor transfer of female households has an incentive effect on farmers’ willingness to quit contracted land, and the effect is more robust than that of incomplete household labor transfer. Seemingly, complete female labor transfer of households has an incentive effect on farmers’ willingness to quit contracted land and the homestead, and the effect is stronger than the complete labor transfer of the family. Because of this, the government should respect the wishes of farmers and strengthen the effective connection and mutual promotion between the homestead and contracted land withdrawal policy. Moreover, pay concentrated attention to the vital role of different types of labor transfer, and targeted labor transfer mechanisms should be used to guide farmers in an orderly manner.

1. Introduction

Homestead farming, also known as subsistence farming, involves small-scale agricultural practices where farmers grow food primarily for their family’s consumption [1,2]. The shift away from homestead farming has been a global phenomenon, especially in regions undergoing rapid urbanization and industrialization [3]. As economies develop, people tend to move away from traditional agricultural practices and migrate to urban areas in search of better economic opportunities [4]. Since the reform and opening up, many rural laborers have moved to cities, and the participation rate of rural laborers in non-agricultural employment has continued to increase, driven by policy incentives and the income gap between urban and rural areas in China [5,6]. According to data from China’s National Bureau of Statistics, the country’s urbanization rate increased dramatically, rising from 17.9% in 1978 to 64.72% in 2021 [7]. The number of migrants in China increased from 6.57 million in 1982 to 281.7 million in 2016 [8]. Seemingly, the total number of migrant workers exceeded 280 million in 2020, and peasants who have entered cities have also begun to migrate to cities and towns in terms of life and residence [9,10]. It should be noted that the transfer of rural labor has also broken the original allocation relationship between family labor and land, making the organic family system of “house, farmland, and people” tend toward fragmenting [11], resulting in severe problems in the allocation and utilization of contracted land and homesteads [12]. The global trend toward reducing homestead farming is not without its challenges. As traditional agricultural practices decline, there are concerns about food security, rural livelihoods, and cultural heritage [13]. Small-scale farmers who rely on subsistence farming may face economic vulnerabilities and limited access to resources and markets. Household labor has moved to cities and non-agricultural industries, and the input and dependence of farm households on contracted land have shown a “double reduction” trend. The trends generally reduce the number of rural consumers or residents and the number of farmers or potential farm labor [14,15].

Moreover, a substantial decrease in the resident population in rural areas increases the construction land and the widespread existence of idle, abandoned, and inefficient use of homesteads and houses [16,17]. The withdrawal of farmers’ land will change the scale of operation of Chinese farms which may exist, and land ultimately needs to be managed [18]. With the transfer of household populations to cities and non-agricultural industries, farmers’ dependence on contracted land and homesteads has decreased. Extensive management or even abandonment of contracted land, idle and inefficient use of homesteads and houses is a widespread problem [19]. Reducing rural homesteading in China leads to significant changes in rural areas, such as a declining population, economic transformation, and environmental impacts [20]. This issue of land resources violates the principle of intensive land use. It causes significant waste of land resources [9,10] and is not conducive to ensuring food security and achieving high-quality development of rural revitalization and urbanization [21,22]. However, rural revitalization efforts can help address these challenges by creating attractive opportunities for the youth, promoting diversified economic activities, and implementing sustainable practices [23]. Through strategic investments in education, technology, and infrastructure, rural revitalization can foster balanced and sustainable rural development, ensuring the prosperity of rural communities and preserving their cultural heritage [24]. Therefore, the issue of farmers’ land withdrawal under the background of labor transfer has become an essential part of the rural land system reform and the “rural revitalization” strategy, which has attracted widespread attention in academic circles.

Farmers’ land withdrawal refers to farmers’ abandonment of land rights such as contracted land and homesteads [25,26]. However, decision-makers face challenges to form the bond between rural people and land by supporting farmers to transfer employment to cities and settle there [27,28]. Many necessary decision-making plans and major strategies at the decision-making level emphasize that the withdrawal of rural land must respect farmers’ wishes and allow them to make their own choices [29]. For this reason, relevant studies (such as Liu et al. [30], Chen et al. [12], and Si et al. [31]) have focused on the willingness of farmers to withdraw from their land and studied that under the background of urbanization and non-agriculture employment. Where farmers have the characteristics of “leaving land without releasing soil” and “releasing soil without releasing rights”, and only a few farmers under the premise of compensation are willing to withdraw from the homestead and contracted land [32,33]. From the linkage of factors, it can be seen that there is a close relationship between the labor force of farmers and the allocation of land factors [34,35].

Moreover, the labor transfer reduces the family agricultural labor force and the resident population in rural areas. However, it is still controversial how it affects farmers’ willingness to withdraw from land [36]. One view holds that, with urbanization and the vigorous development of secondary and tertiary industries, the income gap between agriculture and non-agricultural industries drives rural labor to leave the land and transfer to non-agricultural sectors [37]. The dependence on contracted land and the homestead has increased farmers’ willingness to withdraw from contracted land and the homestead [20,38]. However, some studies (such as Qian et al. [39], Su et al. [40], and ChunLei et al. [41]) found that labor transfer does not necessarily improve farmers’ willingness to withdraw from land, which is reflected in the following two aspects. First, farmers are rational people, and second, they can maximize family benefits through the sensible division of family labor in non-agricultural employment and agricultural production. The transfer does not necessarily lead to the farmer’s land withdrawal; a nonlinear relationship exists between the labor transfer and the farmer’s land withdrawal willingness [42].

The above research has portrayed a crucial research significance. However, the notion of measuring the impacts of labor transfer for withdrawing from farmland has not yet been comprehensively evaluated by any study. Moreover, a combined perspective of contracted and homestead farming has not been explored critically yet. The current study is unique in the following prospectives: First, as the basic unit of rural society, the production and life of peasant households are unified, and the contracted land and the homestead, respectively, carry two complementary functions of production and life. This close connection makes the peasant households simultaneously consider the contracted land withdrawal and the land withdrawal. The homestead is withdrawn to maximize the utility of the family. Existing studies (Gu et al. [29], Tang et al. [17], and Chen et al. [12]) have used single-equation models such as the Logit or the Probit model to study a specific aspect of farmers’ willingness to withdraw from contracted land or the homestead, ignoring the relationship between the two and potential endogeneity issues, and the estimated results may be biased. Second, the existing literature (such as Fan et al. [43], Ouyang et al. [44], and Xu et al. [45]) mainly dealt with the overall non-agricultural income of the family or the number of people transferred and has not made a detailed distinction between the complete labor transfer of “perennial migrant workers” and the incomplete labor transfer of “slack rural migrant workers” within the family, failing to clarify the differences within the family. Different types of labor transfer affect farmers’ willingness to withdraw from farmland [46,47]; therefore, the study includes the heterogeneity of farmers’ types within its core analytical framework [48]. Third, existing studies have not paid much attention to the impact of the female family labor force transfer. The urban–rural income gap and salary mechanism make the dominant male labor force transfer first, and women are the last resort for maintaining the house, taking care of children, and eventually they need to handle the family’s land management [49]. Likewise, to a certain extent, the farmer holds the right to take major decisions on the withdrawal of family land, so the transfer of the female labor force may have a “last straw effect” on farmers’ willingness to withdraw from land.

However, the influence of rural labor transfers on farmers’ willingness to quit contracted land and the homestead has not been comprehensively evaluated. Because of this, based on considering the correlation effect between the withdrawal of contracted land and the withdrawal of the homestead, this study uses the survey data of 953 farmers in Shaanxi, Sichuan, and Anhui to construct a bivariate Probit model that can deal with endogenous problems, and empirically analyses the effect of labor transfer on farmers’ land withdrawal. The unique contribution of this research is its focus on the correlation between the withdrawal of contracted land and homesteads. Examining how labor transfer at various family levels influences farmers’ willingness to withdraw from land enhances our understanding of the root causes of such transitions and their implications in rural areas. The findings offer valuable insights for the government to enhance the design of the rural land withdrawal system and related support policies.

2. Theoretical Analysis and Hypotheses

Farmer household land withdrawal includes contracted land and the homestead, which are essential to rural land system reform. The following discussion depicts the theoretical viewpoints regarding the impact of labor transfer on farmers’ willingness to withdraw from land based on the correlation effect between the two, and proposes corresponding hypotheses.

2.1. Correlation Effect of Farmers’ Willingness to Withdraw from Contracted Land and the Homestead

Farmers are rational economic entities [50], always trying to identify and choose institutional arrangements that benefit them [26,51]. For example, suppose farmers can withdraw from contracted land and homesteads voluntarily. In that case, they will inevitably calculate the potential costs and benefits of various decisions to make the most beneficial choice for the family, whether to “exit” or “not to withdraw” from contracted land and homesteads [34,52]. Based on social experience and facts, it can be seen that, as the basic unit of rural society, the production and livelihood of rural households are essentially interlinked and unified [16,53]. The contracted land and homestead have different natural attributes and undertake different functions to serve the production and life of rural households. The natural attributes of contracted land are production combined with labor to form agricultural output income [54,55]. The natural attribute of homestead land is life, mainly providing safe shelter for farmers, more manifested as an asset function and ethical function [56,57].

Therefore, under the limited institutional arrangement, once farmers can only withdraw from one of the contracted land or homestead, their production and life will be separated, the dividends of family production and life combination will disappear, and some additional factors should be considered [10,33,35]. If a farmer withdraws from the contracted land, he will lose all of the income from the agricultural output and bear the cost of housing maintenance and care. In addition, if farmers can only choose to withdraw from the contracted land refund or homestead, the degree of “separation of the family business” and “living under the fence” of farmers will be exacerbated, making families bear more burdens in livelihood arrangements and emotional maintenance. Emotional maintenance refers to the need for farmers to invest time and money to maintain standard family order. Once the degree of household separation and dependence among farmers increases, emotional maintenance costs, such as long distances and high costs, also increase. It can be assumed that the contracted land and the homestead undertake the essential functions of production and life and have the unity of serving the production and life of the peasant household [58]. Therefore, there is a direct link effect between the willingness to withdraw from the contracted land of the peasant household and the willingness to withdraw from the homestead. Because of this, this paper puts forward Hypothesis H1:

H1. 

There is a direct link effect between the willingness to withdraw from farmers’ contracted land and the willingness to withdraw from homesteads.

2.2. The Impact of Labor Force Transfer on the Willingness to Withdraw from Farmers’ Contracted Land and the Homestead

Farmer’s households are the unity of production and life, and allocating production factors is dynamic. The input quantity and structure of family human and land factors will change with household labor and land factors [29]. Rational farmers will pursue the most optimal value based on the utility maximization of the human–land configuration structure [59]. Along with the unprecedented economic growth and progress over the last 30 years, China has eventually transformed its economy into a non-agricultural economy by flourishing its manufacturing sector and witnessing rapid urbanization trends, which led rural laborers to be poured into cities and towns to seek a better livelihood in secondary and tertiary industries [60,61]. This detrimental effect on the family agricultural labor force and permanent resident population will inevitably lead to the reallocation of household land resources [62]. First, with the gradual transfer of family labor to cities and other industries, the massive loss of family labor has reduced the amount of labor available for agricultural production and shortened agricultural labor hours [63,64]. Dependence on land has increased farmers’ willingness to withdraw from contracted land. Second, with the increase in the number and degree of family labor transfer, the lifestyle and source of livelihood of peasant households are increasingly shifting from agriculture and land, and they gradually settle down in cities, which reduces the demand for peasant households for homesteads and their attached houses [65,66]. As a result, in recent decades, farmers’ willingness to quit homesteads has increased significantly [67]. Some evidence shows that the emotional attachment and link with the root impacts may lead rural residences to carry on the homestead. For example, Xingyu and Yukun [68] and Yan et al. [38] explored the potential reason for keeping homesteads in regional areas of China. They emphasized that, compared with contracted land, homesteads and their attached houses also have crucial emotional value and are the essential carrier of “nostalgia” and attachment to the hometown. In a study of Flanders, Belgium, Rogge and Dessein [69] highlighted that farmers usually do not cut off their connection with their hometown and eventually tend to build up a strong emotional bond with their land and farmstead and keep on farming to be able to stay on their farm. However, it should be emphasized that most of the existing literature (such as Liu et al. [70], Ge et al. [71] and Liu [72]) found in such cases are primarily managed by older adults or women members of the family. Therefore, the transfer of household labor force has a more substantial incentive effect on the willingness to quit the farmer’s contracted land than the willingness to quit the homestead. Based on the above argument, it can be hypothesized that rural labor transfer might have more driving forces to a homestead than other factors, such as emotional attachments. The article put forward the Hypothesis two (H2) as follows:

H2. 

There is a positive impact of labor force transfer on the willingness to withdraw from farmers’ contracted land and homestead.

2.3. The Impact of Different Types of Labor Transfer within the Family on the Willingness to Quit the Farmer’s Contracted Land and the Homestead

Farmers are rational people and pursue the maximization of family effects. They are willing to withdraw from contracted land or the homestead only when the risk is tolerable and the benefits of land restitution are significantly more significant than the cost [73,74]. The transfer of the labor force is the prerequisite for farmers to quit their land, and the willingness of farmers to quit their contracted land and homestead is based on the permanence and stability of household labor transfer [75]. The different degrees of permanence and stability of the process of labor transfer within the family include complete and incomplete parts. Complete labor transfer refers to all family members who have entirely left agriculture and obtained full-time jobs in cities or other industries [76]. Incomplete labor transfer refers to the proportion of family members who have not entirely left agriculture and obtained part-time jobs in cities or other industries [77]. Household incomplete labor transfer has the characteristics of high mobility and “urban-rural migrator” round-trip characteristics, and its non-agricultural livelihood is not stable and long-term [78]. Losing land will significantly increase the vulnerability and potential risks of its livelihood [79], and incomplete transfer of household labor requires both continuing to engage in agricultural production to maximize the value of personal labor. Also, it requires homesteads to guarantee agricultural production [80].

Therefore, the incomplete labor transfer of households has a limited incentive effect on the willingness to withdraw from farmers’ contracted land and homesteads. Compared with the incomplete labor transfer of the family [81], the complete labor transfer of the family represents the family members who have entirely left agricultural production and migrated to urban areas [82]. Their non-agricultural livelihoods are stable and long-term, and they do not need to rely on agricultural production. The value of personal labor can be maximized; purchasing urban housing or stable housing also makes such farmers not need homesteads to protect their housing after the withdrawal of non-agricultural industries [9,83]. Therefore, the complete transfer of the household labor force substantially affects the willingness to quit the farmer’s contracted land and homestead. Given this, this paper proposes Hypothesis H3.

H3. 

There is a positive impact of different types of labor transfer within the family on the willingness to quit the farmer’s contracted land and homestead.

2.4. The Impact of the Female Labor Force Transfer on Farmers’ Willingness to Withdraw from Contracted Land and the Homestead

Land exit is the decision of farmers to maximize the family effect under the division of labor among family members [84]. The number of the family agricultural labor force and the number of rural permanent residents have also declined repeatedly. The female family labor force transfer includes two parts: female complete labor force transfer and female incomplete labor force transfer [85]. Among them, female complete labor force transfer refers to the proportion of all female family members who have entirely left agriculture and obtained full-time employment in cities or other industries. Household women do not complete labor transfer refers to the proportion of all female family members who have not entirely left agriculture and obtained part-time jobs in cities or other industries [86]. With the transfer of the female household labor force to cities and non-agricultural industries, the degree of household de-agriculturalization has increased significantly, and the dependence on contracted land and homesteads has further decreased. What needs to be emphasized is that the complete labor transfer of women in the family and the incomplete labor transfer of the family may have a higher impact on the willingness to quit the farmer’s contracted land and homestead than the complete labor transfer and the incomplete labor transfer of the family, respectively [87].

This is because the transfer of family labor is a gradual process. Male labor with physical strength, technical advantages, and higher market remuneration is the first to transfer to cities and non-agricultural industries. Therefore, practically the family significantly needs to rely on female family members. In China, the trend is popularly known as “Men workers and women farmers”. In this trend, most family land management and other household supports depend on female family members [88]. When rural families initiate female members to relocate to urban areas, the “tight balance” between the family agricultural labor force, resident population, and land allocation might have been broken [5]. The continuity of family land management has been seriously weakened, which led to breaking the “last straw” that ultimately prompted farmers to withdraw from contracted land and homesteads [89]. As a result, the positive effects of complete labor transfer and incomplete labor transfer of female households on farmers’ contracted land and homestead exit willingness are stronger than complete labor transfer and incomplete labor transfer. Because of this, this paper proposes Hypothesis H4. Based on the above discussion, the study proposed the conceptual framework (Figure 1).

H4. 

The female labor force transfer may positively impact farmers’ willingness to withdraw from contracted land and the homestead.

3. Materials and Methods

This study empirically analyses the impact of labor transfer on farmers’ willingness to withdraw from farming (contracted farming and homestead). Therefore, empirically, a bivariate Probit model has been adopted to craft its findings. As the study focuses on two types of farming, a bivariate Probit model would be one of the best options, as it can effectively handle two scenarios in a unified framework. The selected model rectifies a clear advantage over other potential models, such as Logit and Probit. Moreover, it can effectively deal with potential correlation effects and endogenous problems between the two scenarios. Therefore, the current study setup is expected to provide more robust estimations. The paper utilizes STATA 14.0 (StataCorp, Sacramento, CA, USA, www.stata.com) software for regression. The “atrho” value of the bivariate Probit model can be used to test the correlation effect between farmers’ willingness to quit their contracted land and homesteads [90,91,92]. The positive, negative, and significant levels of the coefficients (α and β) can serve as the basis for determining the validity of the hypotheses.

The study refers to existing research to ensure the results of the model reliability, such as Liao et al. [93], Wang et al. [94], and Beck et al. [95]. In addition to the transfer of the labor force, it includes the household characteristics, family characteristics, government, clan, urban–rural gap, and explanatory variables that affect farmers’ willingness to withdraw from land, which can avoid the endogenous problems caused by missing variables to a certain extent and improve the model estimation effect. Moreover, considering the representative relationship between contracted land withdrawal and homestead withdrawal, we choose two groups (i) complete labor transfer and (ii) incomplete labor transfer. Therefore, the analysis framework of this study should be more suitable for farmers’ decision-making scenarios, and the results are relatively more reliable.

3.1. Data Sources

The study used a multiphase or multistage cluster sampling technique to determine the possible respondent. According to the core assumption of the tactics, a probable sample has to be chosen from a diverse population using smaller and smaller groups at each stage [96,97]. While random interviews were used to determine the survey objectives to ensure the data’s validity, the survey method was conducted via face-to-face interviews accompanied by a structured questionnaire. First, according to the agricultural and socio-economic development level and conditions, the study purposively selected Shaanxi, Sichuan, and Anhui provinces as the primary survey areas. After that, the team conducted informal discussions with local agricultural extension officers to choose the most suitable towns and villages according to the context and design of the study. Based on the suggestion obtained by the local agricultural extension officer and the feasibility of the investigation, ten county-level units were determined, including Xixiang, Ziyang, Baihe, Hanbin in Shaanxi, Wangcang, Tongjiang, Mount Emei, and Jinan in Anhui and Qimen, Huangshan in Sichuan province. After that, from each county (city), we randomly selected 2–3 townships, and from each township, we randomly chose 2–4 villages. Finally, we randomly selected 10–15 farmers from each village to collect the empirical data.

The following reasons encourage us to choose these three provinces, obtained from the critical discussion with the local agricultural extension officer and initial feasibility assessment. Firstly, Anhui, Sichuan, and Shaanxi Provinces are distinct regions with higher net population outflows in China. In particular, the net population outflows of the Anhui Province and Sichuan Province are often at the forefront of China. Many of them are transferred from rural agriculture to urban and non-agricultural industries, so the selected study area is representative in terms of labor force transfer. Second, farmers in Anhui, Sichuan, and Shaanxi provinces tend to follow the trends of “leave farming” rather than “leave power”, and “abandon farming” rather than “abandon land”, distorting the relationship between people and land, resulting in extensive management and even abandonment of contracted land. The permanent rural population has decreased significantly, but the construction land has increased rather than decreased. Homestead and housing idle and inefficient use is widespread, so the selected research area is realistic. Third, the Anhui Province, Sichuan Province, and Shaanxi Province are important pilot regions for promoting land withdrawal reform in China. Therefore, according to the value of scientific sampling, selecting farmers from these regions can reveal more applicable laws and provide decision-making references. Figure 2 depicts the study area.

In the formal research stage, the research team (consisting of 12 postgraduate and doctoral students) were uniformly trained by three professors (well-experienced in agriculture economics and management) on the design, contents, and sampling methods of the survey questionnaire. Based on the core design, prior knowledge of the team and review of the existing literature in the field, the study determines all of the important factors and variables and constructs a structured preliminary survey questionnaire. The study conducted a pilot survey among 30 randomly selected farmers (10 from each division) in the Hanzhong, Baoji, and Ankang divisions of Shaanxi Province in early July 2022 to test the questionnaire. In addition to the farmer household questionnaire, each sample asked village cadres to complete the corresponding village-level questionnaire. Based on the preliminary results of the pilot study, this article modified and optimized the final survey questionnaire.

The first choice of research object was the head of household or the primary decision-maker of household production. If the head of household or the primary decision-maker of family production was unavailable, we interviewed other household members with relative experience in farming and over 18. Before interviewing the respondent, the research team, led by the supervisor, visited the village head to eliminate any critical issues in conducting the final investigation. The formal survey was carried out in July to August 2022. A total of 1020 interviews were obtained, and after eliminating the questionnaire with missing critical data, a valid questionnaire set of 953 farmers was obtained, with an effective rate of 93.43%. This could be a significant reason for a high response rate. Moreover, due to some restrictions of COVID-19, the team obtained formal permission from local health administration offices and maintained all of the recommended measures. The distribution and capacity of sample farmers are shown in

Table 1. Research area and sample distribution.

Province	Area	Sample Size	Percentage (%)
Shaanxi Province	Xixiang County	77	8.08
	Ziyang County	107	11.23
	Baihe County	110	11.54
	Hanbin District	80	8.39
Sichuan Province	Wangcang County	91	9.55
	Tongjiang County	99	10.39
	Emeishan City	102	10.70
Anhui Province	Jinzhai County	96	10.07
	Qimen County	94	9.86
	Huangshan District	97	10.18

.

3.2. Variable Selection

3.2.1. Explained Variables

According to the existing studies, the willingness of farmers to withdraw from land includes two aspects: (i) the willingness to withdraw from contracted land and (ii) the willingness to withdraw from the homestead [98,99]. Since the current market price of rural land exit is not yet clear [100], this article adopts the alternative market method to control the impact of price factors. We consider the local average annual land rent as the compensation price for the exit of contracted land and ask the farmers whether they are willing to withdraw from the contracted land. Whereas the price is used as compensation for the homestead’s withdrawal, the farmers are asked whether they are willing to withdraw the homestead.

3.2.2. Key Explanatory Variables

Labor transfer generally refers to the transfer of rural labor to cities and non-agricultural industries [9,101]. According to current trends of urbanization and industrialization, like most other fastly growing emerging economies, rural agricultural practices in china are narrowing. Like most other emerging countries, historically, rural China has been characterized by strong gender norms, where men were primarily responsible for agricultural work, while women were expected to perform domestic and caregiving tasks [102]. This division of labor was reinforced by traditional values emphasizing gender roles and hierarchies, specifically in the rural agricultural labor system. In recent decades, China has undergone significant economic and social changes, which have impacted labor transfer patterns in rural areas and the household labor transfer scenario [103]. Therefore, the degree of permanence and stability of labor transfer between these two groups must be different, and there may also differ in potential livelihood and living arrangements [104]. Likewise, other demands and allocation decisions for household homestead and contracted land are seemingly diverse. Because of this, this study divides labor transfer into two parts (i) complete labor transfer and (ii) incomplete labor transfer [105]. The proportion of full-time workers in cities or other industries and the incomplete labor transfer of households refers to the ratio of all family members who have not wholly separated from agriculture and obtained part-time workers in cities or other industries [106,107]. Five characterization variables of labor transfer were set up from different levels to analyze the impact of labor transfer on farmers’ willingness to withdraw from land. First, the family’s overall labor transfer is represented by “the sum of the family’s incomplete labor transfer and complete labor transfer divided by the total family population”, the sum of labor force divided by the total family population”, and “the sum of the family’s fully transferred labor force divided by the total family population”. Second, for the incomplete labor force transfer of female family members and the complete labor force transfer of female family members, the “sum of female labor force transferred from the family incompletely transferred divided by the total family population” and “sum of the female labor force completely transferred by the family divided by total family population”.

3.2.3. Other Explanatory Variables

Referring to existing research (Such as Qian et al. [39], Zhao et al. [34], and Chen et al. [5]), farmers’ willingness to withdraw from land is not only closely related to household characteristics and family characteristics, but also several other external factors such as the application of intelligent technology, government, clan, and the urban–rural gaps such as better public health and schooling facilities, better livelihood opportunities, and retirement benefits may play a vital role in farmers’ desire to withdraw from farmlands. This study includes the following explanatory variables to ensure the validity of the model and construct a more realistic decision-making scenario for farmers: the age of the head of the household, the education level of the head of the household, the proportion of the elderly population, the sense of land control, the scale of land, the scale of land outflow, total household income, clan status, smartphone use, government interaction, government trust, government obedience, public variables such as enthusiasm for participation in affairs, general trust, evaluation of social policy implementation, evaluation of environmental policy implementation, life satisfaction, perception of urban–rural welfare gap, and rural–urban housing purchase rate. In addition, to control the influence of some unobservable regional factors, the county-level dummy variables are combined with the current county-level governance background. The variable meaning, assignment, and descriptive statistics are shown in

Table 2. Variable meaning and assignment.

Variable	Variable Meaning and Assignment	Mean	Standard Deviation
Homestead Exit Willingness	Unwilling = 0, willing = 1	0.261	0.440
Willingness to withdraw from contracted land	Unwilling = 0, willing = 1	0.301	0.459
Whole family labor transfer	The sum of incomplete labor transfer and complete labor transfer in households/total household population	0.392	0.279
Family incomplete labor transfer	The sum of the incompletely transferred labor force by households/total household population	0.204	0.252
Family complete labor transfer	The sum of the fully transferred labor force of the family/the total population of the family	0.183	0.205
Incomplete labor transfer of family women	The sum of the incomplete labor force transferred by women in the family/total population of the family	0.062	0.130
Household women complete labor transfer	The sum of the total labor force transferred by female households/total household population	0.067	0.117
Age of head of household	(years old)	57.143	10.022
Education level of the head of the household	Illiteracy = 1, primary school = 2, junior high school and technical secondary school = 3, high school = 4, junior college = 5, undergraduate = 5, master’s degree = 6	2.274	0.835
The proportion of the elderly population	The sum of older people over the age of 60 in the family/the total population of the family	0.898	0.895
Sense of land control	Weak sense of land control = 1, average sense of land control = 2, strong sense of land control = 3	1.727	0.547
Land size	(mu)	15.486	20.494
Land outflow scale	(mu)	0.190	1.479
Total household income	(10,000 yuan)	9.663	13.046
Clan	The surname does not belong to the big village surname = 0, and the surname belongs to the big village surname = 1	0.441	0.497
Smartphone use	Very unfamiliar = 1, less familiar = 2, familiar = 3, somewhat familiar = 4, very familiar = 5	2.597	1.322
Government interaction	Very little interaction = 1, less interaction = 2, average interaction = 3, more interaction = 4, less interaction = 5	2.039	1.012
Government trust	The average value of three 5-level scales of trust degree of county government, township government, and village committee	3.454	0.751
Government obedience	Very bad = 1, poor = 2, fair = 3, better = 4, very good = 5	3.889	0.637
Enthusiasm for participation in public affairs	Very low motivation = 1, poor motivation = 2, average motivation = 3, good motivation = 4, very good motivation = 5	3.080	1.085
Universal trust	Most people in society are trustworthy, most people behave fairly and justly, and most of the time, people are helpful	3.835	0.535
Social Policy Implementation Evaluation	The average value of the 5-level scale of rural social policy satisfaction for rural pensions, rural medical insurance, and poverty alleviation and disability assistance	3.726	0.751
Environmental Policy Implementation Evaluation	The average value of the 5-level scale of rural environmental policy satisfaction for the two items of environmental pollution control and beautiful rural construction	3.778	0.732
Life satisfaction	The weighted average of three 5-level scales for quality of life, happiness, and future life	3.611	0.562
Perception of Welfare Gap between Urban and Rural Areas	Compared with the city, the average value of the 2 5 scales of farmer welfare and rural child development	3.404	0.814
Household purchase rate in villages and towns	Very low = 1, low = 2, average = 3, high = 4, very high = 5	2.143	0.928

.

3.3. Key Explanatory Variables

Let $Y_{wc}{}^{*}$ be the latent variable of farmers’ willingness to withdraw from $Y_{wz}{}^{*}$ contracted land, be the latent variable of farmers’ willingness to withdraw from the homestead, $Y_{wc}$ and be the observed variables $Y_{wz}$ of farmers’ willingness to withdraw from contracted land and homestead, respectively. The relationship between latent variables and observed variables in the bivariate Probit model is shown in Equations (1) and (2):

$$Y_{wc}=\{\begin{array}{l}=1,Y_{wc}{}^{*}>0\\ =0,Other=0\end{array},Y_{wz}=\{\begin{array}{l}=1,Y_{wz}{}^{*}>0\\ =0,Other=0\end{array}$$

(1)

$$\{\begin{array}{l}{Y}_{wc}{}^{*}=\alpha X_{wc}+\epsilon \\ Y_{wz}{}^{*}=\beta X_{wz}+\sigma \end{array}$$

(2)

$E(\epsilon )=E(\sigma )$, $Var(\epsilon )=Var(\sigma )$, there is no correlation between the random disturbance term and the explanatory variables. In the formula $X_{wc}$, $X_{wz}$ are the influencing factors of farmers’ willingness to withdraw from contracted land and their willingness to withdraw from the homestead, respectively. Therefore, the probability determination equation of the bivariate Probit model is as follows:

$$P(Y_{wc}=1,Y_{wz}=1)=P(Y_{wc}{}^{*}>0,Y_{wz}{}^{*}>0)={\displaystyle {\int }_{-\infty }^{\alpha X_{wc}}{\displaystyle {\int }_{-\infty }^{\beta X_{wz}}\varphi (w_1,w_2,\rho )}}d{w}_{1}d{w}_2=\Phi (\alpha X_{wc},\beta X_{wz},\rho )$$

(3)

Formula (3) $\varphi (w_1,w_2,\rho )$ and $\Phi (\alpha X_{wc},\beta X_{wz},\rho )$ are the standardized two-dimensional normal distribution probability density function and cumulative distribution function, respectively, and $\rho$ is the correlation coefficient.

Regarding variable processing, labor transfer and each dimension are determined by proportion, which can better reflect the importance of labor transfer to families, as suggested by Wang et al. [108]. Control variables such as general trust, government trust, social policy implementation evaluation, environmental policy implementation evaluation, life satisfaction, and perception of the urban–rural welfare gap were obtained by the arithmetic mean according to the local conditions and the existing literature. Since the standard deviation of related variables is greater than the mean value and more discrete, the total household land area, the scale of land outflow, and the total household income are logarithmic and can be brought into the model substantially [109]. Therefore, based on the above discussion, the data acquisition and variable construction performed by the study are legitimate and can reveal the reality of farmers’ homestead or contracted land withdrawal.

4. Results

A bivariate Probit model was constructed in STATA 14.0 (StataCorp, www.stata.com) to analyze the effect of labor transfer on farmers’ willingness to quit land and the correlation effect of willingness to abandon the land. As suggested by Yu et al. [110], a multicollinearity diagnosis was performed on the variables before the model regression [110]. Whereas VIF < 3 indicates that there is no severe multicollinearity between the variables [111], the robust option is used to deal with the heteroscedasticity problem to ensure the robustness of the model estimation results. The results of the bivariate Probit model in

Table 3. Model results and tests.

Variable	Model 1		Model 2		Model 3
	Willingness to Withdraw from Contracted Land	Homestead Exit Willingness	Willingness to Withdraw from Contracted Land	Homestead Exit Willingness	Willingness to Withdraw from Contracted Land	Homestead Exit Willingness
	Coefficient (Standard Deviation)	Coefficient (Standard Deviation)	Coefficient (Standard Deviation)	Coefficient (Standard Deviation)	Coefficient (Standard Deviation)	Coefficient (Standard Deviation)
Whole family labor transfer	0.512 *** (0.185)	0.241 (0.184)	-	-	-	-
Family incomplete labor transfer	-	-	0.612 *** (0.205)	0.149 (0.205)	-	-
Family complete labor transfer	-	-	0.622 ** (0.269)	0.512 * (0.262)	-	-
Incomplete labor transfer of family women	-	-	-	-	0.737 ** (0.357)	0.186 (0.402)
Household women complete labor transfer	-	-	-	-	0.818 ** (0.408)	1.115 *** (0.395)
Age of head of household	−0.009 (0.006)	−0.006 (0.006)	−0.009 (0.006)	−0.006 (0.006)	−0.010 (0.006)	−0.007 (0.006)
Education level of the head of the household	−0.161 ** (0.064)	−0.082 (0.060)	−0.163 *** (0.063)	−0.079 (0.060)	−0.154 ** (0.063)	−0.078 (0.061)
The proportion of the elderly population	0.181 (0.070)	0.070 (0.071)	0.187 *** (0.069)	0.075 (0.070)	0.153 ** (0.068)	0.068 (0.068)
Sense of land control	0.176 * (0.090)	0.027 (0.092)	0.175 * (0.090)	0.023 (0.093)	0.152 * (0.091)	−0.002 (0.094)
Land size	0.014 (0.071)	0.063 (0.067)	0.019 (0.071)	0.064 (0.067)	0.006 (0.071)	0.066 (0.067)
Land outflow scale	−0.035 (0.153)	0.054 (0.145)	−0.033 (0.153)	0.040 (0.147)	−0.046 (0.149)	0.032 (0.148)
Total household income	−0.037 (0.068)	−0.056 (0.071)	−0.049 (0.073)	−0.090 (0.075)	−0.022 (0.066)	−0.080 (0.072)
Clan	0.221 ** (0.088)	0.221 ** (0.092)	0.227 *** (0.088)	0.223 ** (0.092)	0.226 *** (0.087)	0.226 ** (0.094)
Smartphone use	0.091 ** (0.040)	0.111 *** (0.042)	0.091 ** (0.040)	0.114 *** (0.042)	0.088 ** (0.040)	0.109 *** (0.042)
Government interaction	0.095 * (0.051)	0.185 *** (0.053)	0.102 ** (0.050)	0.185 *** (0.053)	0.084 * (0.050)	0.175 *** (0.053)
Government trust	0.080 (0.078)	−0.062 (0.080)	0.082 (0.078)	−0.060 (0.080)	0.080 (0.079)	−0.056 (0.080)
Government obedience	0.104 (0.079)	0.051 (0.086)	0.099 (0.079)	0.054 (0.087)	0.118 (0.081)	0.054 (0.087)
Enthusiasm for participation in public affairs	−0.338 *** (0.047)	−0.399 ** (0.051)	−0.343 *** (0.048)	−0.397 *** (0.051)	−0.333 *** (0.047)	−0.396 *** (0.051)
Universal trust	0.266 *** (0.098)	0.146 (0.102)	0.261 *** (0.097)	0.154 (0.102)	0.269 *** (0.097)	0.154 (0.102)
Social policy implementation evaluation	−0.031 (0.069)	0.118 (0.073)	−0.032 (0.070)	0.118 (0.073)	−0.039 (0.069)	0.110 (0.072)
Environmental policy implementation evaluation	−0.271 *** (0.075)	−0.285 *** (0.080)	−0.266 *** (0.075)	−0.285 *** (0.080)	−0.265 *** (0.076)	−0.284 *** (0.080)
Life satisfaction	0.102 (0.096)	0.038 (0.098)	0.105 (0.097)	0.049 (0.099)	0.106 (0.097)	0.050 (0.099)
Perception of welfare gap between urban and rural areas	0.109 * (0.059)	0.087 (0.061)	0.111 * (0.059)	0.082 (0.061)	0.111 * (0.059)	0.085 (0.061)
Household purchase rate in villages and towns	-	0.103 (0.081)	-	0.103 (0.080)	-	0.102 (0.079)
Regional variable	Yes		Yes		Yes
Athrho	0.858 *** (0.074)		0.858 *** (0.074)		0.853 *** (0.074)
${\chi }^2$ test (p-value)	134.159 (0.000)		133.605 (0.000)		132.868 (0.000)
Wald chi2	213.88		213.74		221.86
Prob	0.000		0.000		0.000

Note: ***, **, * represent the 1%, 5%, and 10% significant levels, respectively. The total household land area, land outflow scale, and household income are logarithmic and brought into the model; the brackets are robust standard errors.

show that the Wald test ${\chi }^2$ values are all above 130, and the p value = 0.000 passes the 1% significance test, indicating that there is a correlation between the willingness of farmers to quit their contracted land and their willingness to quit their homestead [112]. Therefore, it can be assumed that the Bivariate Probit model is suitable for the data, and the specific results and analysis are as follows.

4.1. The Correlation Effect of Farmers’ Willingness to Withdraw from Land

In Table 3, the “atrho” values of model 1, model 2, and model 3 passed the 1% significance test, indicating that there is a significant correlation effect between farmers’ willingness to withdraw from contracted land and homesteads and a positive coefficient value indicates farmers’ willingness to withdraw from contracted land. In contrast, the correlation effect with the willingness to quit the homestead is complimentary. Therefore, H1 is verified. Farmers are considered rational actors whose decision-making goal is to maximize household utility [113]. Homestead and contracted land are attributed to living standards and production. Combining the two can achieve the effect of “1 + 1 > 2”, maximizing the household utility. If one of the two is missing, farmers will not maximize the utility of their products and support their family needs. For example, only expropriation of a farmer’s homestead for urban construction will cause significant inconvenience to farmers’ resources and arrangement of production activities. Only contracted land is expropriated, so farmers can only enter the city and enter enterprises to find employment opportunities, which eventually brings huge emotional issues and unnecessary expenses. Therefore, a significant direct link exists between farmers’ willingness to quit their contracted land and homestead. Moreover, the correlation between the two situations also effectively explains why the proportion of farmers who are “unwilling to withdraw from contracted land and homestead” and those who are “willing to withdraw from contracted land and homestead” is very high.

4.2. The Correlation Effect of Labour Transfer to Withdraw from Land

The overall labor transfer of households has a positive impact on the willingness of farmers to quit contracted land at the 1% significance level. Still, it has no significant effect on the willingness of farmers to quit their homesteads. The overall family labor transfer consists of two parts: incomplete family labor transfer and complete family labor transfer. However, since the available labor transfer of the family includes insufficient labor, these people must rely on the homestead to meet their housing needs during agricultural production or after exiting the non-agricultural industry, which weakens the pull of the overall family labor transfer to the exit of the farmer’s homestead. Therefore, the general labor transfer of households has no significant effect on the willingness of farmers to quit their homesteads. The transfer of incomplete family labor positively impacts farmers’ willingness to quit contracted land at the 1% significance level. Still, it has no significant impact on the willingness of farmers to quit their homesteads. The transfer of complete and incomplete family labor positively affects the 5% and 10% significance levels, respectively. Conversely, the coefficient comparison shows that the positive effect of household complete labor transfer on farmers’ willingness to quit contracted land and the homestead is more potent than that of incomplete household labor transfer. Therefore, it can be assumed that it affects farmers’ willingness to quit contracted land and the homestead.

The family’s incomplete labor transfer represents the family members who are dissociated between agriculture and non-agricultural, rural and urban. Farmers are more willing to withdraw from contracted land. However, although this part of the people is also engaged in non-agricultural work, they have not entirely separated from agriculture. The family must retain the homestead and attached housing to meet the needs of this part of the people to return to their hometowns; thus, reducing the impact of the incomplete labor transfer of the family on the withdrawal of farmers’ homesteads. Therefore, it has no significant effect on the willingness of farmers to withdraw from their homesteads. The complete family labor transfer represents family members who can rely on non-agricultural livelihoods and maximizes the value of their labor in the non-agricultural industry.

Similarly, households’ reliance on contracted land has been significantly reduced, strengthening farmers’ willingness to withdraw from contracted land. On the other hand, with the complete labor transfer of households, farmers have suitable housing, better wages, and social security in cities or non-agricultural industries, reducing their return to their hometowns. Moreover, coupled with the compensation for the withdrawal from homesteads and the avoidance of rural residential maintenance and depreciation costs, farmers are more willing to withdraw from homesteads. The incomplete labor transfer of female households positively impacts the farmers’ willingness to quit contracted land at the 5% significance level. However, it does not significantly affect their willingness to quit their homesteads. On the other hand, the complete labor transfer of female households has a significant impact at the 5% and 1% significance levels, respectively, on the farmers’ decisions. Thus, it can be assumed that it positively impacts farmers’ willingness to withdraw from contracted land and the homestead. However, from the comparison of coefficients, it can be observed that, when a household has complete or incomplete transfers of their female labor, it significantly impacts farmers’ decisions to withdraw from contracted land and the homestead compared to households where no female labor force transfer takes place.

The mechanism of the complete and incomplete labor transfer of all working women in the family on farmers’ willingness to quit contracted land and the homestead is the same, because separating family labor from agriculture and rural areas reduces farmers’ dependence on contracted land and the homestead. Here, we focus on explaining why the effect coefficients of family female incomplete labor transfer and family female complete labor transfer are larger than those of incomplete and complete family labor transfer, respectively. However, due to the traditional family structure of China, the transfer of the labor force is a gradual process, and traditionally the male labor force is usually the first choice of the other family members. The family division mode of “men’s farming and women’s weaving” has changed to “men’s workers and women’s farmers”, and family agriculture and rural life rely on women. The further transfer of the family female labor force will weaken the family agricultural production, reduce the family rural permanent population, and destroy the existing “tight balance” between the family agricultural labor force, the permanent rural population, and the land allocation. It will become a cause for urging farmers to withdraw from contracted land and homesteads.

4.3. The Effect of Other Explanatory Variables on Farmers’ Willingness to Withdraw from Land

According to the model results in Table 3, the cultural level of household heads has a negative impact on farmers’ willingness to withdraw from contracted land at a 5% significance level. The higher the cultural level, the more farmers value the potential for future appreciation of contracted land. Thus, their willingness to withdraw from contracted land is lower. The educational level of the head of household has a significant negative impact on the willingness of farmers to withdraw from contracted land. The higher the education level, the more peasant households are aware of the outcomes of contracted land, so their willingness to withdraw from the contracted land is low. The sense of land control has a significant positive impact on farmers’ willingness to withdraw from contracted land. The sense of land control is the basis for farmers’ land allocation. From the land belonging to the state to the collective and then to themselves, farmers’ sense of land control is significantly improved, enhancing their willingness to withdraw from contracted land. While the “Clan” status significantly and positively affects farmers’ willingness to withdraw from contracted land and homestead. Farmers whose surnames belong to large surnames in the village have a stronger voice and sense of security and are in a favorable position in land withdrawal, so their willingness to withdraw from contracted land and homestead is stronger. The use of smartphones significantly and positively affects farmers’ willingness to quit their contracted land and homestead. Smartphones help farmers obtain relevant information on contracted land withdrawal, understand its value and significance, and increase farmers’ willingness to withdraw from contracted land.

The awareness of land ownership positively affects farmers’ willingness to withdraw from contracting land at a 10% significance level. The recognition of land ownership is the basis for farmers’ land allocation. From land belonging to the state to belonging to the collective and then belonging to themselves, farmers’ sense of land control has significantly improved, which can enhance their willingness to contract out of the land. The clan situation positively impacts farmers’ willingness to withdraw from contracted land and the homestead at a significance level of 5%. Farmers with surnames belonging to larger surnames in villages have more substantial discourse power and a sense of security and are in a favorable position in land withdrawal. Therefore, their willingness to contract and homestead land withdrawal is more vital. The use of smartphones has a positive impact on farmers’ willingness to withdraw from contracted land and the homestead at a significance level of 5%. Smartphones help farmers obtain relevant information about contracted land and its withdrawal, recognize its value and significance, and enhance their willingness to withdraw and withdraw from contracted land.

The government interaction positively impacts farmers’ willingness to withdraw from contracted land and the homestead at a 10% significance level. Government interaction shapes farmers’ sense of policy identity by transmitting policy information related to land withdrawal and enhancing farmers’ willingness to withdraw from contracted land and the homestead. The enthusiasm for participating in public affairs has a negative impact on farmers’ willingness to withdraw from contracted land and the homestead at a significance level of 1%. The withdrawal of land means that the community foundation for farmers’ public participation is dissipated, so farmers with high enthusiasm for public affairs participation are less willing to contract land and homestead land withdrawal. Universal trust positively affects farmers’ willingness to withdraw from contracted land at a significance level of 1%. General trust can reduce farmers’ expectations of the difficulty and risk of land withdrawal. Thus, farmers with high general trust are more willing to contract the land. The evaluation of environmental policy implementation has a negative impact on farmers’ willingness to withdraw from contracted land and homestead at a significance level of 1%. Environmental policies have improved rural areas’ natural environment and infrastructure, strengthening farmers’ confidence in agricultural and rural development. Therefore, farmers who believe environmental policies are well implemented are less willing to withdraw from contracted land and homesteads. The perception of the urban–rural welfare gap significantly impacts farmers’ willingness to withdraw from contracted land at a 10% significance level. The perception of the urban–rural welfare gap has a significant attraction on farmers’ land withdrawal. When farmers perceive a large urban–rural development gap, they give up contracted land and go to cities to make a living, increasing their willingness to withdraw from contracted land.

5. Discussion

Rapid urbanization and the consequent transfer of labor from rural areas are two influential factors currently redefining the landscape of homesteading practices on a global scale [114]. The expansion of urban areas and the shift towards industrialization have led to significant changes in homesteading practices, as people are increasingly moving away from traditional rural lifestyles. Due to the swift urbanization process, the accessibility of extensive rural territories to establish homesteads has significantly diminished [115]. As urbanization continues to expand and influence rural regions, agriculturally productive parcels that were once well-suited for establishing homesteads are being supplanted by urban landscapes dominated by concrete infrastructure. The limited availability of appropriate land presents a growing difficulty for individuals aspiring to become homesteaders, as they encounter obstacles in locating and establishing their self-sustaining retreats [116].

Furthermore, the attraction of urban living and the availability of employment prospects in emerging sectors results in a substantial migration of the rural workforce to urban areas. There is a growing trend among young adults and families to migrate toward urban areas for enhanced employment opportunities, improved educational resources, and convenient access to contemporary amenities [117]. Rural-to-urban migration exacerbates the diminishing number of individuals inclined towards homesteading, as they embrace urban lifestyles instead. Moreover, the transition from a rural to an urban setting entails alterations in lifestyle choices and the accessibility of resources. Traditional homesteading is predicated upon utilizing natural resources, including copious amounts of water, fertile soil, and expansive areas suitable for agricultural practices and animal husbandry [118].

Nevertheless, swift urbanization can give rise to various limitations in terms of resources, such as restricted availability of uncontaminated water, pollution of urban soil, and spatial constraints that pose challenges for traditional homesteading methods [119]. Specifically, the trends are significantly visible within emerging economies, as most thrive on transiting their economy and gradually shift from subsistence-based farming to large-scale industrialized framing [120]. Moreover, massive transitions are also occurring in their manufacturing sectors to cope with global and local demands for diverse products [121]. However, such transitions of leaving homesteads have become widespread from sub-Saharan Africa throughout Latin America to south and southeast Asia [122].

Seemingly, urbanization and labor migration are well-connected phenomena in China, primarily powered by profound economic development trends [123], which eventually create more opportunities for rural labor forces to move to urban regions, to seek better lifestyles and income opportunities [124]. The trends became more visible after “the reform and the opening up” policy was endorsed in the 1970s [125]. This approach increased the rural-to-urban gap by encouraging a business economic recovery that minimized agriculture participants and emphasized non-agricultural business industries [71]. Due to several structural restructurings implemented over the last three decades, the cultural and social components of the agricultural sector have undergone significant transformations. This transformation includes the social obligation, farmers’ attitudes and perceptions regarding the homestead, social and economic inequality, and government development policies’ effectiveness [126]. However, the existing literature does not fully grasp whether labor transfer improves farmers’ willingness to withdraw from farming. Thus, the study aims to comprehensively explore the area to evaluate the notion of labor migration and withdrawal from agriculture.

The descriptive statistics of the sample farmers depict that 61.805% of the farmers are unwilling to withdraw from the contracted land or the homestead, 12.067% of the farmers are only willing to withdraw from the contracted land, 8.080% of the farmers are only willing to withdraw from the homestead, and 18.048% of the farmers are willing to withdraw from the contracted land and exit the homestead. This also effectively explains why the proportion of farmers who are “ unwilling to withdraw from the contracted land nor the homestead” and those who are “ willing to withdraw from the contracted land and homestead “ is relatively high. The comprehensive assessment of the study shows that, overall, the willingness of farmers to withdraw from contracted land and the homestead is not high, which is well aligned with the study of Zhao et al. [34] and Fan and Zhang [11]. Interestingly, the study outlined that there is a significant direct link effect between the willingness of farmers to withdraw from contracted land and the willingness to withdraw from the homestead. There is a direct link effect between farmers’ willingness to withdraw from contracted land and their willingness to withdraw from the homestead. The outcomes are parallel with the assumption by Si et al. [31]. However, in a study of Jinhu County, Jiangsu Province, China, Zhao et al. [34] found distinctive outcomes. They articulated that eliminating homesteads is essential in increasing the effectiveness of rural development land usage and a practical strategy to improve rural farmers’ lifestyles.

Likewise, different levels of labor transfer affect farmers’ willingness to withdraw from contracted land and homesteads. The overall labor transfer of households can improve farmers’ willingness to withdraw from contracted land. The findings are supported by the study of Ayinde et al. [127]. But, it does not significantly affect farmers’ willingness to withdraw from homesteads. Carte et al. [128] found different results in a study of Nicaragua and Guatemala. They depicted that the interviewed farmers have shown relatively higher perceptions of leaving the homestead due to increasing overall labor transfer. The study also concluded that incomplete labor transfer of households could improve farmers’ willingness to quit contracted land. Still, it does not significantly affect their willingness to quit their homesteads. The findings are also supported by the study of Cai and Ng [129] and Asfaw et al. [130]. Likewise, complete family labor transfer positively affects farmers’ willingness to quit contracted land and the homestead, and the effect is more robust than households without complete labor transfer. The incomplete labor transfer of female households significantly impacts farmers’ willingness to quit contracted land, and the effect is stronger than that of incomplete household labor transfer. The results are well in parallel with the assumption found by Leibbrandt et al. [131]. In contrast, the complete transfer of the female labor force of households has a significant impact on the willingness of farmers to quit contracted land and the willingness of contracting teams to leave, and the effect is stronger than the complete labor transfer of the family. The outcomes are well aligned with the study of Peel et al. [132], Carter-Leal et al. [133], and Hansen [134].

The above discussion depicts that the government should highlight the critical factor associated with strengthening the effective connection and mutual promotion between the exit policy of the homestead and contracted land. They should take the labor transfer assessment as an essential assessment indicator to promote the reform of the contracted land and homestead system and guide farmers with a high degree of family labor transfer, to prioritize exiting the contracted land and the homestead in an orderly manner. Interestingly, the current construction of the rural homestead and contracted land use rights system in China is not perfect, many regulations are unclear, and there are even legislative conflicts [5]. There is a lack of institutional protection for farmers’ land-related rights, such as compensation for contracted farmers when land is expropriated and requisitioned, which cannot be directly reflected. Additionally, there is a lack of a long-term mechanism for compensating landless farmers, a lack of transferability and collateral value, and vogue land speculation and hoarding, which may limit the betterment of land rights in rural areas. There are many principles and rules regarding the acquisition and change in the right to use rural homesteads, but they have few specific regulations. However, the Chinese government has acknowledged these challenges and has implemented measures to reform and enhance the land use rights system. Initiatives have been undertaken to tackle concerns about compensation, transferability, and property rights clarity, among other matters. The implementation of extensive reforms in a vast and heterogeneous nation such as China necessitates a considerable amount of time, while the deficiencies within the system persist and remain a subject of deliberation and examination among policymakers and scholars. Therefore, the Land Administration Law should more comprehensively and strictly regulate the procedures for obtaining the right to use rural homesteads. It is necessary to strengthen the relevant legislation on the circulation form, circulation condition, circulation procedure and supervision procedure of rural homestead use rights, and continuously refine it to reduce the disputes arising in the transaction process of rural homestead use rights.

Moreover, the scope of the original subject of rural homesteads should be comprehended. The government should formulate the rural collective economic organization according to the actual situation, the focus on work, life, living conditions, and other factors of the parties. In addition, it is necessary to strengthen the guarantee of the economic attributes of rural homesteads; such as, reasonably standardize the transaction price of the right to use the rural homestead, improve the role and status of the market and the transaction subject, rather than the government, in the transaction process. They should protect farmers’ interests, fully respect farmers’ will, and improve rural residents’ participation in the withdrawal compensation of rural homestead use rights. At the same time, the government should undertake the relevant work of law popularization, strengthen publicity, improve the legal awareness of farmers, and help them to understand the relevant professional knowledge of rural homestead use of right withdrawal mechanism, to enhance the ability of rural residents and to protect their rights and interests. Likewise, promoting the optimization of family labor transfer should be strengthened as an essential driving force to facilitate the betterment of the potential reform. Under-improvised marginal farmers, especially female farmers, should be prioritized in the reform as they could be most economically vulnerable. Vocational skills training and agriculture industrialization should be facilitated within rural regions, which could be crucial for creating more jobs in rural areas.

Agricultural industrialization may be pivotal in retaining labor in rural areas and is closely linked to China’s “rural revitalization” concept [135]. To maintain a balance between rural and urban migration, the line between “Agricultural Modernization and Industrialization” and “Rural Revitalization” should also be strengthened. By modernizing farming practices and introducing advanced technologies, agricultural industrialization enhances productivity and efficiency, leading to increased agricultural incomes [136]. Creating non-agricultural job opportunities in agribusinesses and agricultural processing industries reduces the push factors for rural laborers to migrate to urban centers for better employment prospects [137]. This labor retention in rural areas aligns with China’s broader goal of rural revitalization, which seeks to bridge the development gap between urban and rural regions [138]. The government should extend its support to create vibrant and sustainable rural communities, fostering balanced regional growth and social development by promoting rural development, improving living conditions, and integrating agriculture into modern supply chains.

Interestingly, it is essential to consider the reasons for keeping homestead farming rather than just keeping the property [56]. Land preservation is valuable, but differentiating it via homestead farming has additional advantages. Homestead farming encourages self-sufficiency by enabling people to generate food and resources, minimizing the reliance on outside sources. Additionally, it supports environmentally sound and resilient agriculture techniques, including organic farming and agroforestry [139]. Maintaining homestead farming fosters a feeling of place, protects cultural heritage, and strengthens local economies. One may develop a strong connection with nature and ensure biodiversity protection by actively interacting with the land via homesteading. Therefore, considering this difference can potentially build resilient, self-sufficient communities firmly entrenched in the environment [59]. While retaining land holds inherent value, the specific practice of contracted land farming presents distinct advantages [16]. Maintaining contracted land farming enables cultivating specialized crops or livestock, promoting enhanced efficiency and productivity. Leasing arrangements facilitate farmers in directing their attention towards their specialized areas of expertise and gaining access to larger parcels of land [76]. By maintaining ownership of contracted land for agricultural purposes, individuals can take advantage of economies of scale, facilitate resource sharing, and mitigate operational expenses [60]. Furthermore, this methodology promotes the cultivation of cooperation and alliances within the agricultural sector, thereby generating prospects for exchanging knowledge and emerging innovative practices [73]. Future researchers should evaluate these potential factors in their core analytical framework.

6. Conclusions

The study’s main objectives are to focus on whether the labor transfer can effectively affect the farmers’ willingness to withdraw from the homestead and contracted land and whether there are degree differences and gender differences in this impact. Therefore, by considering the correlation effect of contracted land withdrawal and homestead withdrawal, we focus on the differential impact of labor transfer and its different types within farmers household’s land withdrawal behavior. Moreover, the study utilizes the notion of labor transfer, type differences, gender differences and farmers’ willingness to withdraw from the land to test the impact of labor transfer on farmers’ willingness to withdraw from land and measure the correlation effect between the willingness to withdraw from contracted land and the homestead. The theoretical baseline of the study has been set based on the existing literature and the bivariate Probit model. In contrast, the empirical data have been compromised with the survey data of 953 farmers collected from the Shaanxi, Sichuan, and Anhui provinces.

Through these analyses, the impact mechanism of labor transfer on the willingness to withdraw from contracted land and the homestead is to reduce farmers’ dependence on the land, thus affecting farmers’ willingness to withdraw from contracted land and the homestead. The “atrho” values of Model 1, Model 2, and Model 3 show a significant correlation between farmers’ willingness to contract land and their willingness to withdraw from homesteads. A positive coefficient value indicates a complimentary correlation, confirming hypothesis one (H1). The overall labor transfer of households has a positive impact on farmers’ willingness to withdraw from contracted land at a 1% significance level, but no significant impact on their willingness to withdraw from homesteads. The complete transfer of household labor force has a significant positive impact on farmers’ willingness to withdraw from contracted land and homestead at 5% and 10% levels, respectively. The positive effects of incomplete labor force transfer from female households and complete labor force transfer from female households on farmers’ willingness to contract land and homestead land exit are higher than those of incomplete labor force transfer from households and complete labor force transfer from households. Assuming H4 is verified. Overall, the findings suggest that the relationship between farmers’ willingness to contract land and their willingness to withdraw from homesteads is complex and multifaceted. This paper discusses the following policy implications based on the above results.

(i)

Because of the current low willingness of farmers to withdraw from contracted land and homesteads, the government should, on the one hand, steadily make forward-looking institutional arrangements for land withdrawal to lay the institutional foundation for farmers to withdraw from contracted land and homesteads in a gradual and orderly manner. On the one hand, the government must have historical patience and responsibility for the withdrawal of farmers’ contracted land and homesteads, respect the wishes of farmers, and resolutely prevent farmers from being “excluded from the land”.

(ii)

Therefore, the willingness to withdraw from the homestead needs to be premised on the willingness to withdraw from contracted land, or the willingness to withdraw from contracted land can strengthen the willingness to withdraw from the homestead. As the correlation effect between farmers’ willingness to withdraw from contracted land and their willingness to withdraw from the homestead, the two should be placed within the framework of the rural land withdrawal system, and the practical connection and mutual promotion of the homestead withdrawal policy and the contracted land withdrawal policy should be strengthened. Farmers voluntarily withdraw from contracted land as the guide, establish a policy support system that supports the effective combination of contracted land withdrawal and homestead withdrawal, and guide farmers to withdraw from contracted land and the homestead in an orderly manner.

(iii)

The government should provide farmers with more non-agricultural jobs through the secondary and tertiary industries, improve the vocational skills training to enhance the competitiveness of farmers ‘ non-agricultural employment and promote the development of family labor from incomplete transfer to complete the transfer. On the other hand, the government should take the complete labor transfer of households as a reference to improve the targeting of the land exit policy, improve the land exit policy supported by employment resettlement, and use “industry” to return “land” to improve the exit of farmers’ contracted land and homestead will.

(iv)

Since the prominent role of female family labor transfers on farmers’ willingness to quit contracted land and the homestead, the government should increase the skills training for rural women’s off-farm employment and promote participation. At the same time, it should be committed to creating a better job environment for rural women. A more equitable and improved non-agricultural employment environment will advance in the long-term and stable transfer of the female labor force, reduce farmers’ dependence on contracted land and homesteads, and increase their willingness to withdraw from contracted land and homesteads.

In this study, farmers’ willingness to withdraw from contracted land and the homestead are included in the unified analysis framework while focusing on the current context of the Chinese rural labor transfer mechanism. However, the trend of leaving farmland has become more evident in several emerging countries and even in developed nations. Therefore, the notion should gain much more attention globally. Thus, the framework of the study should be utilized in different parts of the world to grasp more diverse and fruitful outcomes. On the contrary, due to its design, the study mainly focused on farmer’s household surveys; there are enough possibilities to contain biased responses. The research team discussed all the variables and essential information with the respondent to reduce the issue. However, labor transfer may usually foster by various interpersonal and socio-demographic factors, which could have impacted the mechanism diversely. Seemingly, the extent of the right to use rural homestead by law could be a significant property interest for farmers. Thus in future research, these aspects should be highlighted and included in core assumptions based on the framework we provided. The law can limit the disposal of the homestead by the owner, but the right to gain should also be endowed and protected to protect farmers’ interests.

The study focused on only three provinces, potentially not representative of whole countries’ scenarios. Therefore, future studies should choose different regions and types of farmers and obtain more meaningful conclusions. In addition, due to the lack of enough tool variables, adopting endogenous processing methods other than the bivariate Probit model is challenging. So potential studies should utilize structural model-building tactics such as Structural Equation Modelling (SEM) and Interpretive Structure Modelling (ISM). Future studies can consider expanding the implication and types of labor transfer, including more classified variables, building more realistic decision-making scenarios, and profoundly revealing the impact of labor transfer on farmers’ willingness to withdraw from land. Moreover, they can consider integrating labor transfer with farmers’ land withdrawal intention and land withdrawal behavior and introducing Multivariate Probit for in-depth analysis to achieve innovation in analysis framework and model application. In contrast, the impacts of rural revitalization by empowering rural modernization and industrialization should be included in the core study framework to obtain more robust outcomes.