**2. Literature Review**

In domestic and foreign scholars' studies on land transfer, the relevant influencing factors are mainly investigated in three sequential dimensions: before-during-after. Regarding the "before land transfer" scenario, the studies mainly focus on the impact of individual endowment differences of farmers or households in different regions on willingness to transfer land or the area of the transfer or transfer out deadline. Regarding the "during land transfer" scenario, the studies mainly focus on the consideration of the transaction method or the form of land transfer. Regarding the "after land transfer" scenario, the studies mainly focus on the impact of land transfers on the changes in the welfare of the transferee.

Many scholars have conducted relevant research on the impact of family, regional, and individual endowment differences on land transfer willingness, decision-making, or transfer area. In terms of family endowment differences, the employment situation of family members is one of the important factors affecting the willingness to transfer land in rural areas. Research shows that if a family member can obtain a stable non-agricultural job in urban areas, it will promote the decision-making of family land transfer, and the member will have the greatest power in decision-making [14–17]. Further, the stronger the willingness of farmers to seek non-agricultural jobs in cities, the more inclined they are to long-term land transfer [18]. At the same time, the family income structure also has an impact on land transfer. The higher the proportion of migrant workers' labor income in the total household income, the more likely farmers are to choose land transfer or abandonment [19]. From the perspective of regional differences, scholars used a panel dataset of 171 Chinese cities that developed high-speed rail infrastructure from 2005 to 2012 and applied the SEM model to find that the expansion of the high-speed rail network had a significant impact on the circulation of agricultural land, and the impact of high-speed rail on the circulation of agricultural land in the western region is five times that in the

eastern region [20]. Specific to the differences in various rural location factors, the study found that rural site resources have a significant impact on the circulation of agricultural land. Communities with good infrastructure, that are close to towns, with sufficient labor force, and with high economic input and output do not rely on agricultural land, but natural conditions and well-connected communities rely more on agriculture [21]. From the perspective of individual endowment differences among farmers, the level of financial knowledge possessed by farmers obviously affects land circulation, and financial literacy has a greater impact on land inflow than outflow [22]. Based on the survey data of 8031 households with 27 identities in China in 2014, scholars found that the experience of famine in the early years impacted land circulation [23]. In addition, the older the household head is, the stronger the attachment to the land, and the less likely the household head is to transfer land [24]. In addition, the conclusions of academic research are relatively consistent where the degree of risk perception and attitude, the land property rights system, and the degree of awareness of land policy also significantly affect farmers' willingness to transfer land [25–27].

Secondly, with consideration of the form of land transformation transaction or land transferred out, some scholars have found that land cooperatives promote farmers' land transfer [28]. Meanwhile, some scholars found that when comparing the land stock cooperation mode with rural cooperatives as the main body, against the land leaseback and re-contract mode with "village collective + planting leading firm" as the primary management body, the improvement of farmers' livelihood capital was more evident [29].

Finally, in terms of the impact on the welfare of land transferees after transferring, one study found that the welfare effect of the transferred-out farmers was higher than that of the transferred-in farmers [30]. In contrast, one research shows that the subjective welfare of farmers who transfer farmland will not increase. Further analysis shows that the subjective welfare of farmers who trade with acquaintances is higher than that of farmers who trade with non-acquaintances [31].

In summary, most previous studies have focused on the factors influencing willingness in each link before-during-after land transfer, the transaction transfer mode, and the subsequent net welfare value; especially, the literature focusing on farmers' willingness to transfer their land based on their endowment characteristics is fruitful and has formed a useful reference for subsequent studies. However, there is limited literature on the effect of heterogeneity of endowments on the area of land transferred out based on different household structures, and this is where this paper is expected to make a marginal contribution.

## **3. Materials and Methods**

#### *3.1. Research Hypothesis*

Traditionally, family structures mainly include nuclear, joint, main, single-parent, and broken families, and different family structures have prominent heterogeneity in resource endowment. Different household structures related to different household life cycles and different household life cycles influence the relevant decision of the family, such as entrepreneurship, land transfer, land scale operation, household consumption, and labor supply [32–35]. Further, households are divided into elite households and handicapped households, and ordinary households. Generally, elite households are relatively rich in various resources. They tend to have more advantages in household decision-making. Still, they then may capture some national policy dividends and then become "elite capture," and "elite capture" exists mainly in resource allocation and precise poverty alleviation in rural areas of China [36–38]. For other households, especially broken families, the existence of "elite capture" may lead to a more serious policy deviation. Based on the above principles, this paper focuses on whether the heterogeneity of family structure affects land transfer area from the actual land transfer area. Moreover, this paper divides rural households into elite, broken, ordinary households based on existing studies, and focuses on the different influences of elite family and broken family on land transfer, and then explores whether

there is a "crowding out effect" in China due to the heterogeneity of family structure in the size of the land transferred out. Based on relevant studies, this paper proposes the following hypotheses.

**Hypothesis 1.** *Elite households have an advantage in land transfer and are more likely to transfer land.*

In this paper, the rural elite households are defined as political elite households, divided into households with party members and households with grassroots cadres. There is little literature on the effect of the two households on the difference in their land transfer area. The existing studies generally generalize the two households as political capital. However, in real life, for party members and grassroots cadres in rural areas, there are indeed some differences between them. Party members (non-grassroots cadres) in rural areas are mostly engaged in non-agricultural work and have a looser attachment to the land, so they are able to grasp the policy dividend of land transfer and are more likely to transfer their family land for rent. For grassroots cadres, their main workplaces are in rural areas, and they are more or less engaged in agriculture-related work and more closely connected with the land. Based on the status, hypotheses 2 and 3 are proposed as follows.

**Hypothesis 2.** *Party households have an advantage in land transfer and may transfer a larger land area.*

**Hypothesis 3.** *The advantage of grassroots cadres' families in land transfer is not obvious, and they are more closely connected with the land, which has no significant effect on land transfer area.*

#### *3.2. Data Resource*

This paper uses data from the 2014 survey of the China Household Income Project (CHIP). In July and August 2014, the China family income project (CHIP) conducted the fifth survey. As the main information collected in the survey is related to the income and expenditure in 2013, it is named CHIP 2013, which is consistent with the previous four surveys. This survey is supported by the National Natural Science Foundation of China and the National Bureau of Statistics (NBS) and organized by the China Income Distribution Institute of Beijing Normal University. The survey was conducted by the National Bureau of statistics. The CHIP 2013 sample is from the annual integrated household survey conducted by the National Bureau of statistics in 2013, which includes 160,000 households in 31 provinces. These samples were screened in the eastern, central, and western regions by a systematic sampling method, involving 15 provinces, 126 cities, 234 counties, 18,948 households, 64,777 people, 7175 urban households, 11,013 rural households, and 760 peasant households. CHIP is considered as one of the best public data sources on household income and expenditure in China [39].

All in all, CHIP 2013 is a nationally representative rural household registration survey data sample, and the data are the latest data source of the database. In addition, in the past three years, studies have shown that under the background of China's vigorous implementation of targeted Poverty Alleviation Policies, the phenomenon of "political elites being captured" still exists in rural areas [40]. Therefore, it can be reasonably speculated that most of the rural areas in China have the national conditions of "elite capture". Therefore, the data of CHIP 2013 used in this paper are timely and representative, and also conform to the current situation of rural areas in China.
