*3.2. Heterogeneity in the Impact of Land Transfer on Household Poverty Vulnerability across Different Types of Farm Households*

The fundamental situation, economic source, and capital composition of the family have a substantial bearing on the impact of land transfer. Farmers experience the most economic pressure and the least social pressure; farmers have the highest adaptability to social capital and the lowest adaptability to financial capital [61]. Cultivated land serves as a social security function for farmers, and land assets not only represent agricultural production means in rural China, but also provide employment security for farmers and can be used as collateral for loans [62], allowing farmers to apply for bank loans [63]. With the improvement of the farmland transfer market, the availability of non-agricultural labor has increased, the average non-agricultural labor time of rural households has risen, and the average household income has increased [64]. The family characteristics of farmers, such as income, sources of income, and social capital, have a direct effect on the likelihood of future poverty [65,66]. Different types of farm households were grouped and regressed in clusters to assess the impact of the land transfer on their VEP.

#### 3.2.1. Clustering Farmer Households According to Their Poverty Level

According to World Bank poverty thresholds of USD 1.9 and USD 3.1 per capita per day, households with per capita daily consumption of USD 1.9 or less are classified as extremely poor, those with a per capita daily consumption of USD 1.9 to USD 3.1 are classified as relatively poor, and those with a per capita daily consumption of USD 3.1 or more are classified as non-poor. Consequently, the impact of the land transfer on VEP of three types of farm households, namely absolutely poor, relatively poor, and non-poor, was investigated. In Table 3, columns 1 to 3 display the regression results of land transfer on VEP of absolute poverty, relative poverty, and non-poverty households. The number of absolute poverty in the sample is 0, so column 1 is empty. The results indicated the following: (1) Land transfer can substantially reduce the household VEP of non-poor and relatively poor farmers. This suggests that non-poor and relatively poor farmers can use land transfer to revitalize their household assets to increase their income, reduce poverty, and improve their ability to resist poverty in the future. (2) The further comparison revealed that the greater absolute value of the estimated coefficient of land transfer on VEP of households with relative poverty suggests that land transfer is more effective at preventing the occurrence of poverty in this population. Relatively poor households are even more lacking in the original accumulation of capital to escape poverty. The poverty trap theory proposes that the vicious cycle of low income–low savings–low capital formation–low output–low income is the root cause of poor households' inability to escape poverty [67]. The short-term rental income effect brought by land transfer can provide relatively poor households with the necessary capital accumulation to break out of the poverty trap.


**Table 3.** Regression results grouped by household poverty level.

3.2.2. Clustering by Criteria of Farm Household Financing Constraints and Government Subsidies

It is common for farmers in rural areas to be constrained by financing, which has become a significant factor impeding their ability to increase their income and reduce poverty. Farmers are divided into financing-constrained households and non-financingconstrained households based on the results of the questionnaire item "Have you ever been denied a loan or credit?" On this basis, the effect of land transfer on the poverty and vulnerability of farmers with varying financial constraints is investigated. Government subsidies refer primarily to whether or not farmers have received various cash or in-kind subsidies from the government, such as subsistence allowances, subsidies for returning farmland to forests, and agricultural subsidies. Government subsidies are an essential policy instrument for alleviating poverty. Farmers were classified as government-subsidized and non-government-subsidized farmers, and the disparate effects of land transfer on the poverty and vulnerability of these two groups were investigated.

The regression results in Table 4 indicate the following: (1) Land transfer has no significant effect on the VEP of rural households with financing constraints, but it has a significant inhibitory effect on the VEP of households without financing constraints. Possible causes include a rise in the demand for credit due to the diversification of farmers' sources of income away from agriculture after the transfer of their land, particularly in the form of non-farm self-employment activities. Non-financing-constrained households can more easily meet credit needs and quickly convert household income to non-agricultural operations, whereas this is difficult for financing-constrained households. (2) Land transfer can effectively reduce the VEP of government-subsidized households while having no significant effect on households that have not received government subsidies. The government's planned provision of targeted vocational education, property development and alleviation programs, and other preferential policies may be the cause of this smooth resolution of the employment issue facing rural land transfer households.


**Table 4.** Regression results grouped by financing constraints and government subsidies.

#### *3.3. Heterogeneity in Head of Household Characteristics*

The individual circumstances of the household's primary breadwinner are a significant factor in that household's level of financial well-being and, by extension, its degree of independence from subsistence vulnerability and external risks, and they vary widely [68]. The job category of the head of household will lead to a widening of the poverty and income

gap, and the impact of human capital on household economic growth is significantly greater than that of physical capital [69]. When farmers retire or become too old to work in agriculture, they will sublease their cultivated land in order to maintain their daily consumption [70]. According to the results of a survey conducted by Glauben et al. [71], educational attainment plays a significant role in reducing poverty. The vulnerability of structural poverty is the primary source of vulnerability in rural China, and compulsory education has significantly decreased the vulnerability to poverty [72]. Universal primary education reduces structural poverty vulnerability more effectively than temporary poverty vulnerability [73]. Further analysis proves that compulsory education primarily improves various abilities to obtain a permanent income, such as cognitive ability and participation in non-agricultural work, which are crucial means of reducing structural poverty [74]. Consequently, it is essential to identify these characteristics that influence household head income and incorporate them into the assessment of land transfer effects, as we will do in this paper.

The relationship between land transfer and a family's VEP will vary depending on the nature of work of the head of household, who is typically the primary decision-maker and backbone of the family. Farmers are classified into three categories based on the employment status of the household's head: self-employed farmers, farmers with unstable employment, and farmers with stable employment. Farmers whose primary source of income comes from their own agricultural operations are considered self-employed, farmers whose primary source of income comes from both agriculture and non-agricultural jobs are considered unstably employed, and farmers whose primary source of income comes from non-agricultural jobs with a stable employer are considered employed farmers. Accordingly, the variability of the impact on farm household VEP with different employment choices of the household head is examined. The regression results are shown in Table 5.

**Table 5.** Regression results grouped by the nature of work of the household head.


Table 5 presents regression results for the effect of land transfer on household VEP for each of the three household head employment choices. Land transfer can significantly reduce the VEP of self-employed farmers but has no significant effect on the VEP of employed farmers. Possible explanations include the fact that the family income structure of self-employed farmers is single, the transfer of farm land creates contiguous production that increases the efficiency of agricultural output, and the income increase effect is obvious, thereby effectively reducing VEP.

#### *3.4. PSM Robustness Analysis*

PSM was developed to test the validity of the preceding conclusions. The farmer households whose farmland was transferred out were assigned to the disposal group, while those who did not participate in the farmland transfer were assigned to the control group. The disposal group and the control group were matched based on their propensity scores, and a balance test was conducted to ensure that there were no significant differences in their main characteristics. The causal relationship between land transfer and VEP of rural households was then investigated. At the same time, based on the family poverty level, financing constraints, government subsidies, and type of work of the household

head as criteria, the subsamples are divided for group testing and the average elimination effect is calculated using two matching methods: nuclear matching and nearest neighbor matching. The results of the PSM robustness test are shown in Table 6.

**Table 6.** PSM method robustness test results.


Note: The average treatment effect on the treated(ATT) is a participant average treatment effect.

Table 6 demonstrates that the ATT of land transfer on VEP of rural households under the two matching methods is negative and statistically significant. In general, land transfer reduces the VEP of farm households by 2.7% to 3.8%. A possible explanation is that China's per capita arable land is small, and the corresponding land transfer area for farm households is small, so the measured land transfer does not have a significant impact on VEP. The ATT in the western region is significantly negative, and land transfer can reduce the VEP of rural households in the western region by about 3.7%. The average disposal effect in the central region is negative but not statistically significant, which is significantly inconsistent with the negative regression coefficients in Table 2, indicating that the effect of land transfer on the VEP of rural households in the central region requires additional investigation. The average disposition effect in the eastern region is also not significant, which partially corroborates the robustness of the regression results in Table 2.

The ATT of absolute and relatively poor households is not statistically significant, which is consistent with the conclusion of the regression analysis presented in Table 3. The average disposal effect of non-poor households is notably negative, and land transfer can reduce the VEP of non-poor households by 3.5% to 3.7%. This verifies the validity of the regression results presented in Table 3.

From the standpoint of financing constraints and government subsidies, the average disposal effect of households with financing constraints is insignificant, whereas the ATT of households without financing constraints is significantly negative. The transfer of land will reduce the VEP of households without financial constraints by 3.6% to 5.4%. The ATT of farmers who receive government subsidies is significantly negative, and land transfer can reduce the VEP of government-subsidized families by 4%, whereas the average disposal effect of farmers who do not receive government subsidies is positive. The effect is inconsequential. This verifies the validity of the regression results in Table 4.

Based on the household head's job nature grouping, the ATT is significantly negative for self-employed farmers, and land transfer can significantly reduce the poverty risk of self-employed farmers by 2.7%. However, the ATT does not have a significant impact on reducing the VEP of the two types of farm households. The robustness of the regression conclusions presented in Table 5 is evident.

In conclusion, most of the research conclusions presented in Tables 2–5 have passed the robustness test.

#### **4. Discussion, Conclusions, and Implications**

#### *4.1. Discussion*

Transferring rural land is conducive to enhancing the utilization efficiency of rural land resources, boosting the competitiveness and comprehensive economic benefits of agriculture, and ensuring the continuous increase in farmers' income. However, land transfer compensates agricultural households primarily for the production function of the land, but insufficiently for its security function and asset function [75]. Farmers who transfer land may become "uncultivated land, insecure, and unemployable" if they engage in new labor, participate in labor market competition [76,77], and adapt to urban life [78]. We believe that the issue of the impact of the land transfer on VEP cannot be generalized and needs to be discussed according to the heterogeneity of farm households.

According to the poverty trap theory, regional differences, material resources, educational level, social capital, and financial constraints all have an effect on poverty [79]. In terms of regional heterogeneity, the higher the level of regional economic development and the more comprehensive the infrastructure, the more conducive they are to reducing the likelihood of future poverty in the region; otherwise, the region will fall into persistent poverty [80,81]. This also applies to the effect of land transfer on reducing poverty. Heterogeneity in family characteristics, family income, financial constraints, government subsidies, and other characteristics will result in greater income and labor dividend heterogeneity resulting from land transfer [82–84]. There are disparities in family wealth, and the primary factor is the household head [85]. For the majority of peasant families, the household head is the determining factor in the family's income [86]. The education and occupation of the household head have a strong explanatory power for the family's wealth [87–89]. According to the findings of this study, land transfer can encourage the diversification of farmers' livelihood strategies, thereby increasing farmers' income. Land transfer facilitates large-scale agricultural production and is a crucial means for farmers to increase their income. The study portrayed the group of farmer households able to reduce VPE through land transfer.

Compared with existing studies, this paper uses the VEP indicator to measure the future poverty risk faced by farming households, overcoming the fact that current studies only measure the welfare level of individuals or households at a certain point in time; it takes into account the heterogeneity of farming households and explores the poverty reduction effect of land transfer on households with different characteristics separately; the use of the PSM method effectively solves the endogeneity problem of VEP.

In addition, there are some shortcomings in this research, which future research would need to address: (1) The modest reduction in poverty vulnerability in the PSM test results may be due to the small size of the transferred land area, which needs to be focused on in the next surveys and studies. (2) The effect of land transfer on the VEP of rural households in the east and central region has not yet been conclusively demonstrated, and further research is necessary to analyze the effects of land transfer on poverty reduction in different regions, disaggregated by level of economic development and conditions of intensive land use. (3) Due to a lack of data, the selection of factors influencing VEP is not comprehensive enough, for example, the distance between the village and the main town can be used as a variable to measure how land values affect the VEP. In the future, we will conduct in-depth research on the aforementioned topics in an effort to arrive at more meaningful conclusions and offer more instructive recommendations for practice.

#### *4.2. Conclusions and Implications*

This paper conducted a categorical regression of farm households through a multidimensional perspective and applied the PSM method to test the effect of land transfer on farm household VEP and its heterogeneity. The findings indicate the following: (1) In rural China, land transfer has been shown to have abatement effects on VEP of between 2.9% and 4.2% for farm households. (2) In the western region, land transfer significantly reduces the VEP of farm households, while in the eastern and central regions, no such reduction

is seen, and the effect on the VEP of farm households in the central region remains to be demonstrated. (3) In terms of household characteristics subgroups, land transfer has no significant effect on the VEP of farming households already in poverty (absolute and relative poverty), while it can significantly reduce the VEP of non-poor farming households. (4) Land transfer was effective in promoting VEP reduction among farm households without financial constraints and with government subsidies, but had no effect on VEP among farm households with financing constraints and without government subsidies. (5) When considering the subgroup of characteristics associated with household heads, land transfer primarily has a significant abating effect on the VEP of self-employed farmers, while having no significant impact on the VEP of employed farmers. These results provide empirical evidence for government land management, agricultural development, and the decisions of farmers.

The preceding findings have significant policy ramifications: The first step is to expedite the process of reforming the system that governs rural land and then to standardize and methodically guide the flow of rural land. Currently, the small amount of arable land per person in China's rural areas, the difficulty of structural adjustment, and the high cost of agricultural production are significant factors limiting the efficiency of agriculture, the increase in farmer income, and the revitalization of rural areas. Transferring rural land is conducive to enhancing the efficacy of rural land resource utilization, boosting the competitiveness and comprehensive economic benefits of agriculture, and ensuring farmers' income growth is sustainable. In addition, the loss of land management rights does not result in severe poverty shocks for farming households, as land transfer enables farmlandtransferring households to harvest land transfer rents, boosts the capital accumulation of farming households, encourages the diversification of farmers' occupations, and enhances the structure of farming households' household income. Secondly, rural land transfer must adhere to the local, village-based, and household-based principles. Due to the varying natural conditions in different regions and the unbalanced rural economic development that has resulted in significant differences among villages and farmers, the promotion of land transfer should be tailored to local conditions, meaning that it should be different from village to village and different from household to household. Furthermore, it should not be carried out in a manner that is too hasty, as this would infringe on the lawful rights and interests of farmers. The third phase is to put in place a system to facilitate the transfer of land in rural areas. The government should propose a series of targeted financial and human capital service protection policies, such as financial assistance for eligible new agricultural businesses to take precedence in agriculture-related projects. Furthermore, issues such as a lack of funds for relocating families and achieving diversification of livelihood strategies and sustainable livelihoods should be addressed. In the meantime, the government should offer targeted vocational and special skills training in non-farming fields to increase farmers' employability outside of agriculture.

**Author Contributions:** Conceptualization, Z.W.; methodology, Z.W., Y.L. and M.Y.; software, M.Y. and Z.W.; validation, Z.W. and M.Y.; formal analysis, Z.W. and Z.Z.; investigation, Z.W.; resources, M.Y.; data curation, M.Y. and Y.L.; writing—original draft preparation, Z.W. and M.Y.; writing—review and editing, M.Y., C.W., Y.L. and Z.Z.; visualization, M.Y. and Z.W.; supervision, M.Y. and C.W. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was financially supported by the National Social Science Foundation of China (Grant No. 20&ZD095), Pan Jiahua Expert Workstation (Grant No. 2021GZZH01), Chongqing Social Science Planning Project (Grant No. 2020ZDSC07, 2020ZDGL07), Chongqing Municipal Education Commission Science and Technology Plan Project (Grant No. KJZD-K20210080), Shanxi Provincial Philosophy and Social Science Planning Project (Grant No. 2021YY04), Chongqing Postgraduate Research Innovation Project (Grant No. CYB22271), and Humanities and Social Sciences Research project of Chongqing Municipal Commission of Education (Grant No. 21SKGH346).

**Institutional Review Board Statement:** Not applicable.

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** Not applicable.

**Acknowledgments:** The authors thank the Institute of Social Science Survey for providing data.

**Conflicts of Interest:** The authors declare no conflict of interest.
