Research Articles

Farmland use behavior and its influencing factors from the perspective of population migration

  • MA Libang , 1, 2, 3 ,
  • ZONG Yanling 1 ,
  • WANG Xiang 1 ,
  • SHI Zhihao 1 ,
  • ZHANG Wenbo 1
  • 1. College of Geography and Environmental Science, Northwest Normal University, Lanzhou 730000, China
  • 2. Key Laboratory of Resource Environment and Sustainable Development of Oasis, Gansu Province, Lanzhou 730070, China
  • 3. Institute of Urban and Rural Development and Collaborative Governance of Northwest, Lanzhou 730070, China

Ma Libang (1984‒), Professor, specialized in rural settlement and its spatial reconstruction. E-mail:

Received date: 2023-02-20

  Accepted date: 2023-09-07

  Online published: 2024-04-24

Supported by

National Natural Science Foundation of China(42271222)

Natural Science Foundation of Gansu Province(22JR5RA130)

Science and Technology Program of Gansu Province(22JR5RA136)


Due to rapid socio-economic development and urban-rural integration, the rural population is increasingly moving away from the primary industry, leading to a noticeable transformation of rural farmland use. This paper analyzed the changes in farmland use and the mechanisms in 213 villages of Longxi county, China in 2020, using multiple linear regression models, based on the aforementioned situation. Analysis revealed main findings: (1) Male and young and middle-aged emigration levels are concentrated in areas with higher and lower values, the emigration of talent is more evenly distributed. Overall, male emigration rates were high in all directions and low in the central area, whereas the young and middle-aged emigration rates exhibited a pattern of high loss in the north and south extremes, and low loss in the central area. The emigration of talent demonstrated a pattern of low losses to the east and high losses to the west of the G30 national highway. (2) Primary farmland use behavior was self-cultivation, then abandonment and finally transfer, with over 60% of the total area in 179 villages used for self-cultivation compared to less than 5% in 164 villages allocated for transfer, while less than 30% of the area in 179 villages was abandoned. (3) Significant differences were observed in the factors that influenced various farmland use behaviors, emigration of male, young and middle-aged and talent were the common influencing factors observed among all three types of farmland uses. The loss of males, young and middle-aged had a significant association with reduced self-cultivation while the emigration of talent led to an increase in self-cultivation use. Increased emigration of all three population constituencies significantly increased farmland transfer and abandonment. The conclusions carry significant theoretical and practical implications for enhancing the coordination of rural human-land relationships and improvement of the understanding of the relationship between population migration and farmland use.

Cite this article

MA Libang , ZONG Yanling , WANG Xiang , SHI Zhihao , ZHANG Wenbo . Farmland use behavior and its influencing factors from the perspective of population migration[J]. Journal of Geographical Sciences, 2024 , 34(3) : 439 -458 . DOI: 10.1007/s11442-024-2212-4

1 Introduction

Ensuring food security is a significant foundation for national rejuvenation, as well as a crucial concern for achieving the Sustainable Development Goals of the United Nations. However, as the COVID-19 pandemic spreads globally, the production per unit area of the three main crops has fallen for the first time in 20 years. This has resulted in almost 3.722 billion individuals confronting food insecurity in just two years (Li et al., 2023). This demonstrates that food security needs to receive adequate attention from around the world. Farmland is a fundamental resource in maintaining global food security (Wang et al., 2022). It serves as the central factor of production for farmers, and provides important functions like social security. However, during the 20th century, the phenomenon of farmland abandonment, triggered by the migration of agricultural workers, declining net land returns, land system reforms and agricultural policy adjustments, was seen everywhere in the world (Ramankutty et al., 2002). As a more effective way of allocating land elements, farmland transfer is a necessary strategy to achieve moderate-scale operations in agriculture (Zhao et al., 2021), but the defects in the property rights system and other problems inhibit its implementation, resulting in suboptimal farmland utilization. Simultaneously, with the continuing spread of urbanization and industrialization, a considerable number of rural laborers have been transferred from the agricultural sector to the non-agricultural sector, and, by the end of 2020, the total number of Chinese migrant workers had reached 285 million (NBSC, 2020). In the context of this mass migration, the workforce involved in agricultural production consists mainly of women and the elderly, resulting in the “feminization” and “greying” of agricultural production (Dong et al., 2022). Reflecting these trends, according to China's third agricultural census in 2016, 47.5% of the labor force in agriculture was female, and 33.6% were over 55 years of age. Therefore, with the goal of strictly adhering to the red line of 1.8 billion mu of farmland to ensure food security, and amidst the backdrop of rural laborers joining non-agricultural employment in significant numbers, there is an urgent need to reverse the waste of high-quality farmland resources, while seeking a good human-land relationship.
Numerous empirical studies have been conducted on the behavior of farmland use, providing rich data covering various regions and countries. Scholars from overseas have analyzed the effects of farmland use behavior and its mechanisms. They suggest that a more comprehensive evaluation of the present status of land transfer, transfer efficiency, income generation, and reducing poverty effects of land transfer is vital in underdeveloped regions like Asia, Africa, and Latin America (Benjamin and Brandt, 2002; Kozak, 2003; Deininger and Jin, 2008; Jin and Jayne, 2013; Huy et al., 2016). Some researchers found that land transfer increased farmers’ income and also had a poverty-reducing effect (Benjamin and Brandt, 2002; Deininger et al., 2008). Farmland use behavior dynamics have been analyzed by some academics from the viewpoint of farmland management and ownership (Rogers et al., 2021). Other studies examined the causes of changes in farmland use in different regions. Industrialization and urbanization are considered the primary reasons for farmland use change in developed countries, including Japan and Europe (Kozak, 2003; Benayas et al., 2007; Mou et al., 2022). In contrast, the use behaviors of farmland in Eastern Europe, such as Ukraine and Poland, is more heavily influenced by land systems, land reform policies, or the limited development of agricultural technology (Baumann et al., 2011; Alcantara et al., 2013). Chinese academics, however, have focused their research perspective on changes in farmland use behaviors in the context of growing industrialization and urbanization and increasing population demand for food. Some academics have analyzed the factors influencing varied farmland usage behaviors and their spatiotemporal evolution characteristics, primarily involving farmland quality, geographical location, transaction costs, family and village characteristics among other factors (Jansuwan and Zander, 2021). Other academics conducted research from the perspective of the relationship between farmland use behaviors and found that the transfer of farmland reduces the probability of abandonment (Shao et al., 2015). Subsequently, some academics developed simulations and models of farmland use behavior (Song et al., 2018), further increasing the range of research methods employed. As China's urbanization continues to advance, the imbalance in the relationship between people and land caused by the massive migration of rural labor has become the primary factor impeding the development of rural transformation (Xu et al., 2019; Wang et al., 2022; Xie and Huang, 2022), which shows that non-farm labor has a significant influence on farmland use behaviors and confirms that the impact changes with the degree of interaction variables. In conclusion, despite academics have carried out numerous studies into the behavior of farmland use, there are still certain gaps to be addressed. First, academic studies on the behavior of the use of farmland have primarily concentrated on one behavior, without acknowledging the existence of multiple other behaviors prevalent in rural areas. Second, most of the current research has centralized around farm households, provinces or national levels, with fewer investigations carried out in villages. Moreover, there is an inadequate amount of investigation on the behavior of farmland use from a labor force structure viewpoint. The relationship between farm households and cropland is multifaceted, and different demographic elements have significant effects on different farmland use behaviors.
Based on this, this paper, from the perspective of population migration, based on the theory of farmer's behavior, and based on the micro survey data of 213 villages in Longxi county, analyzed the characteristics of migrating population and farmland use behavior at the village level, as well as the influence of migrating population on their farmland use behavior, and then explored the action mechanism and differences of farmland use behaviors from the perspective of population migration. And propose relevant policy recommendations. Since previous studies mainly concentrate on one kind of farmland use behavior, in this context, this paper mainly discusses the influencing factors and mechanisms of three kinds of rural farmland use behaviors (self-cultivation, transfer and abandonment), and makes a comprehensive analysis of farmland use behavior on the basis of migrating males, young adults and talents, taking into account factors such as the economy, farmland and location.
As the main body of agricultural production and direct users of farmland, rural labor force's decision-making behavior directly affects crop output. Moreover, due to its large outflow, more fertilizers and pesticides are put into agricultural production to ensure the balance of food supply and demand, thus affecting food security and food safety. Therefore, the main purpose of this study is to analyze the mechanism influencing farmland use behavior from the perspective of population migration. Based on this analysis, reasonable policy recommendations will be presented to regulate farmland use behaviors in rural areas. The rational use of farmland is a crucial guarantee and premise for food security and food safety. It reduces the waste of farmland resources and promotes efficient farmland grain productivity that contributes to sustainable agricultural growth and the resolution of the food supply and demand contradiction. Moreover, it stabilizes farmers’ incomes. In addition, it is beneficial to the formulation and implementation of the farmland protection policy of the western government and provides a reference for the scientific formulation of policy opinions in other regions of China.

2 The theoretical basis

The rural area is an open spatial system with certain structures (Wu et al., 2022), functions and inter-regional connections, with interconnections and interactions involving all internal elements (Liu, 2018). Farmland is a fundamental resource for the survival and development of human society, and population is the main driver of rural development. As urbanization expands, the limited scope for rural labor to boost their income causes a widening gap between urban and rural wages. Consequently, numerous young and strong rural laborers migrate to the cities and towns. This results in the older and weaker people in the country representing an increased proportion of the rural labor force, which in turn leads to the decay of human capital, decreased quality of the workforce and other problems, accompanied by the continuous development of the land transfer market and the deepening problem of abandonment of farmland (Zhu and Cai, 2016). The migration of labor actors between rural and urban areas during the process of urbanization and the response of actors bring about changes in the behavior of farmland use in geographical space (Liao et al., 2021). The relationship between the two variables (labor quality and farmland usage) is studied in the current paper, because of the increasing incidence of rough management or abandonment of farmland.
In the context of the continuing rural exodus, the impact of labor on farmland-use decisions is mainly reflected in two aspects. First, it is the reduction in the agricultural labor force; although moderate labor loss can slow down the labor-intensive arable farming, excessive loss results in inadequate labor available for agricultural production, so that farmland cannot be cultivated (Li et al., 2022), which leads to farmland abandonment and the marginalization of agricultural land (Wang et al., 2014). Second, the process of population migration also reduces the overall quality of the labor force which is available to be engaged in agricultural production. Under the urban-rural push-pull effect, numerous young and able laborers move to the cities to engage in non-agricultural industries, leading to changes in the rural population structure, manifested as an uncoordinated development in terms of gender balance, aging, physical traits, and so on. Thus, the decision with regard to how to use farmland will result in a shift from relying mainly on labor to relying on technical means, such as farm machinery and pesticides. Furthermore, as rational economic agents, farmers’ decisions with regard to agricultural production inputs follow the principle of utility maximization under specific cost constraints (Dong et al., 2022), so they will make appropriate decisions on the use of farmland based on the changes in labor factors. When faced with the fact that the income from farming is inadequate to meet their own living expenses, the motivation of agricultural labor force to engage in agricultural production is undermined. Consequently, they tend to exit the primary sector and turn to non-agricultural industries in an attempt to obtain the maximum economic benefits with the lowest human capital, so as to achieve the maximum satisfaction of personal utility.
Moreover, farmland use behavior is influenced not only by the population, but also by other aspects.
(1) The resource endowment characteristics of farmland are characterized by farmland fragmentation, and the degree of farmland fragmentation affects input costs (Wang et al., 2020), which in turn increases the likelihood of farmers abandoning their land.
(2) Village conditions are characterized by village surplus labor. With sufficient surplus labor in the village, the size of the labor force affects the scale of agricultural labor factor inputs (Wang and Zhao, 2022), which, in turn, affects decisions made by farm households on farmland use.
(3) Economic conditions are mainly characterized by household non-farm income. Higher non-farm income results in a greater proportion of labor input being directed toward non-farm industries (Zhang et al., 2014), with farmers becoming more inclined to abandon their land or to transfer it.
(4) Locational conditions are characterized mainly by the distance to the nearest highway. The distance to the nearest highway indicates the level of accessibility, which further influences the farmers’ decision to use the farmland.
To conclude, the behavior of farmland use is influenced by a combination of these factors.
In summary, farmers will make rational decisions based on the status of production factors, such as household labor, capital, land and market changes. To begin with, the significant emigration of labor has caused various structural changes in it, such as decreasing quantity, increased age, and decreased quality of the labor force. Consequently, farmers are more prone to abandoning farmlands or transferring ownership. Secondly, income from non-agricultural industries is significantly higher than that from agriculture. Due to this imbalance, farmers resort to reallocating their labor to enhance their living standards and gain higher income. As a result, the labor force will increasingly move to non-agricultural sectors, resulting in a change in the frequency of self-cultivation, transfer and abandonment of farmland. Hence, population mobility will undoubtedly have a great impact on farmland use behavior (Figure 1).
Figure 1 Theoretical framework

3 Methodology

3.1 Study area

Longxi county is located in central Dingxi city, the southeast of Gansu province, China. It lies between 104°19′-105°14′E and 34°50′-35°24′N, 52 km wide from east to west and 46 km long from north to south, with a total area of 2,408 km2 and is part of the Loess Plateau. The topography is high in the northwest and low in the southeast, with elevations ranging from 1552 to 2727 m. It is a typical loess hill and gully area (Figure 2). The Wei River crosses from west to east, resulting in three narrow strips at Nanshan, Chengchuan and Beishan, with low and moderate hills in the south, the Wei River valley plain in the center, and the loess hills and gullies in the north. The complex terrain conditions in Longxi county led to a diversification of farmland use behavior, with the proportion of self-cultivated land in 179 villages reaching up to 60%, and the highest being 99.13%. Overall, the proportion of transfer farmland was relatively low. In 49 villages, the proportion was more than 5%, and the highest was only 53.38%. Farmland was abandoned in 213 villages, with the highest abandoned area being 81.91%. At the same time, Longxi county is part of the ancient Silk Road and is also an important material exchange and distribution point in central and southern Gansu. The Lianyungang-Lanzhou (Longhai) Railway runs longitudinally from east to west, the Baolan two lines across the territory, and the main G30 highway runs through the territory, with a number of provincial highways and railways intersecting at Wenfeng town, a “dry dock”, which has the development and diffusion effect of expanding from south to north and connecting east to west. Transport is convenient and the location is superior, making it easy for the labor force to go out for work. Longxi county now has 12 towns and five townships, 213 administrative villages, 11 communities and 1287 village groups. In 2020, the county's total household population was 524,800 people, of which 433,000 people worked in agriculture; to lose 137,300 people from the rural labor pool, accounting for 31.71% of the agricultural population; Specifically, the proportion of male emigration in 40 villages was more than 50%, up to 98.61%; in 62 villages, the proportion of young and middle-aged emigration was more than 50%, and the highest was 98.28%. The proportion of talented emigrants is over 50% in 82 villages, the highest is 98.09%. This large-scale emigration is serious.
Figure 2 Location of the study area (Longxi county, Dingxi city, Gansu province, China)

3.2 Data source

The data in this paper were mainly derived from two sources: (1) 2020 village status survey data: based on Participatory Rural Appraisal (PRA), in April-May 2021, the research team visited Longxi county for a period of 20 days of village research, involving the 213 villages, collecting data on population, social security and facilities; and (2) 2020 land survey data: township and administrative village boundaries and farmland data of Longxi county were obtained from the Longxi County Natural Resources Bureau.

3.3 Research method

3.3.1 Pre-processing of indicators

In the analysis of the factors influencing the use of farmland, it was necessary to standardize the initial data in order to eliminate the influence of the scale and the size of the value on the results. When the larger indicator value had a greater influence on the farmland use behavior, the positive indicator equation, Eqn. (1), was used for standardization; when the smaller indicator value had a greater influence on the farmland use behavior, the negative indicator equation, Eqn. (2), was used for standardization. The calculation equations are as follows:
$Z_{i}=\left[C_{i}-\min \left(C_{i}\right)\right] /\left[\max \left(C_{i}\right)-\min \left(C_{i}\right)\right]$
$Z_{i}=\left[\max \left(C_{i}\right)-C_{i}\right] /\left[\max \left(C_{i}\right)-\min \left(C_{i}\right)\right]$
where Ci and Zi are the original and standardized values of the i-th indicator for each village in Longxi county, max(Ci) and min(Ci) are the maximum and minimum values of the i-th indicator, respectively.

3.3.2 Multiple linear regression model

At the village scale, the model setup for the impact of population mobility on farmland use behavior is as follows:
$\mathrm{Y}=\beta_{0}+\beta_{1} \times M M+\beta_{2} \times Y M+\beta_{3} \times E d u+\alpha_{1} X_{1}+\alpha_{2} X_{2}+\alpha_{3} X_{3}+\alpha_{4} X_{4}+\varepsilon_{i}$
where Y denotes the proportion of various uses of farmland area (self-cultivation, transfer and abandonment), MM denotes male emigration, YM denotes young and middle-aged emigration, edu denotes talent emigration, X1 denotes village characteristics, including village surplus labor, X2 denotes village economic characteristics, mainly in terms of average household non-farm income, X3 denotes village location characteristics, including distance from the nearest main road, X4 denotes plot characteristics, characterized by fragmentation of the farmland, β0, β1, β2, β3, α1, α2, α3, and α4 for model each intercept, εi as random perturbation terms.

3.4 Variable determination and description

Explanatory variables. From the village perspective, the farmland use of the 213 villages in Longxi county mainly includes three aspects: self-cultivation, transfer and abandonment. The proportion of the area of each farmland use type, relative to the total farmland area, is as follows:
$Y=X / S$
where Y indicates the proportion of each type of farmland use (namely self-cultivated, transferred or abandoned), X indicates the area of self-cultivated, transferred or abandoned land, and S indicates the total area of farmland in each village.
Core explanatory variables. Values were calculated for male emigration (the number of males as a proportion of males residing in the villages), young and middle-aged emigration (the number of migrating individuals aged 15-65 as a proportion of the residential population in that age group), and talent emigration (the number of migrating individuals with high school education or above as a proportion of the residential population with this level of education).
Control variables. As there are many factors affecting land use, this paper also considers other factors based on population-related variables. (1) village conditions: the labor force is always the direct user of farmland, so this paper introduces the proportion of surplus labor force in villages to characterize its influence on farmland use; (2) location conditions: good location means having high accessibility, which provides convenience for farmers, and this variable is the distance of villages from the nearest road (Mou et al., 2022); (3) economic conditions: income is the basic guarantee of farm households’ quality of life, and it also affects the decisions of farmland use, so the average non-farm income of households is used (Guo et al., 2020); and (4) farmland conditions: the degree of farmland fragmentation directly affects the efficiency and effectiveness with which agricultural production can be carried out, and most existing studies use univariate variables such as the number of plots, average plot size, or average household area as indicators of the degree of farmland fragmentation. Many researchers also use Simpson's index as a measure of the degree of fragmentation of cropland, which, to a certain extent, makes up for the shortcomings of using univariate variables to measure the degree of fragmentation of cropland. Simpson's index is:
$S_{i}=1-\sum_{i=1}^{n} a_{i}^{2} /\left(\sum_{i=1}^{n} a_{i}\right)^{2}$
when Simpson index Si=0, it indicates that there is only one plot, and the degree of cultivated land fragmentation is the least;when Si=1, it indicates that cultivated land fragmentation is very serious, and the larger the Si, the greater the degree of cultivated land fragmentation; ai indicates the area of the i-th plot of the village; n indicates the total number of plots. The explanation and descriptive statistics of each variable are shown in Figure 3.
Figure 3 Variable selection

4 Results

4.1 Population migration characteristics

Based on the population migration index system and related data, ArcGIS 10.4 was used to classify the results of migration of males, young adults and highly educated individuals into four levels: low, lower, higher and high, and to map the spatial distribution of emigration of males, young adults and highly educated individuals accordingly (Figure 4).
Figure 4 Population migration characteristics of Longxi county, Gansu province, China

4.1.1 The degree of male emigration

The degree of male emigration from Longxi county showed large differences in both quantity and space (Figure 4a). Quantitatively, the degree of male emigration is greatest in higher-value and lower-value areas, with 63 higher-value villages and 86 lower-value villages accounting for 29.58% and 40.38% of the total number of villages, respectively. Spatially, the degree of male emigration was relatively high in the villages at the north and south ends of the county, whereas male emigration was relatively low in the intermediate villages.

4.1.2 Degree of young and middle-aged emigration

Young and middle-aged emigration also showed large differences in both quantity and space (Figure 4b). In terms of quantity, the loss of young adults was mainly in the higher- and lower-value villages, of which 48 were higher value areas and 91 were lower value areas, accounting for 22.54% and 42.73% of the total number of villages, respectively. Spatially, the overall layout is high in the four sides and low in the middle, emigration was high at the north and south ends and low in villages in the intermediate area.

4.1.3 Emigration degree of highly educated people

The emigration rate of highly educated people is shown in Figure 4c. Quantitatively, the number of villages where highly educated people emigrated was relatively evenly distributed, with 43 villages in the high-value area, 51 in the higher-value area, 68 in the lower-value area, and 51 in the low-value area, accounting for 20.19%, 23.94%, 31.92% and 23.94% of the total number of villages, respectively. Spatially, the loss of educated people from villages distributed in the western part of the G30 national highway in the county was high, whereas that from villages distributed in the eastern part of the G30 was low.

4.2 Differences in farmland use

Overall, there were significant differences in farmland use behaviors in Longxi county (Figure 5). Among them, in terms of self-cultivation behavior, the self-tillage area of 179 villages accounted for more than 60%, with an average ratio of 89.44%. Such villages could account for 84.04% of the total evaluation units, and the number of self-tillage villages with less than 60% accounted for only 15.96% of the evaluation units. Therefore, the farmland utilization in Longxi county is still dominated by self-cultivation (Figure 5a). In terms of transfer behavior, the transfer area of 164 villages accounted for less than 5%, with an average ratio of 2.94%. Such villages could account for 77.00% of the evaluation units, and only 49 villages with an area ratio higher than 5% had a relatively low transfer of farmland in Longxi county (Figure 5b). In terms of abandoned farmland behaviour, 179 villages had abandoned farmland area accounting for less than 30%, with an average ratio of 6.29%. Such villages could account for 84.04% of the evaluation units, and 34 villages with a ratio higher than 30% accounted for 15.96% of the total evaluation units. It can be seen that the phenomenon of abandoned farmland is relatively common in Longxi county. There was a relatively low degree of abandonment in each village (Figure 5c).
Figure 5 Farmland use status

4.3 Effect of population migration on farmland use behaviors

4.3.1 Testing of the model

Regression analysis was carried out using Stata software. Before reporting the results of the regression analysis, it was necessary to test the covariance of the explanatory variables needed to be tested first. The mean value of the variance inflation factor was 1.102, as determined using SPSS software, and the maximum value did not exceed 2, indicating that the covariance problem of the explanatory variables could be ignored; to verify the robustness of the results, model (1) was used to perform the regression analysis of the core explanatory variables, followed by model (2) to Model (5), which gradually added control variables for the analysis. In this process, none of the directions of the core explanatory variables changed, indicating that the results were robust. Based on the regression analysis of the different agricultural land uses with their influencing factors using Stata software, models (1)-(5) are the regression results obtained with the stepwise addition of control variables.

4.3.2 Effects of population migration on self-farming behavior

The results (Table 1) showed that the factors significantly influencing self-cultivation were male emigration, young and middle-aged emigration, educated people emigration, surplus labor, non- agricultural income, fragmentation of farmland and distance to the nearest road. Among them, the relationship with male emigration was negatively significant at the 5% level, indicating that male emigration inhibited farmers’ decision to farm, while young and middle-aged emigration and emigration of educated people were both negatively significant at the 10% level. The proportion of surplus labor (positive relationship) and the fragmentation of farmland (negative relationship) were both significantly related to self-cultivation at the 1% level, indicating that both were important factors influencing the decision to engage in self-cultivation, albeit in different directions. Both the non-farm income and the distance to the nearest road positively influenced the decision to farm one's own land, at the 10% and 1% levels respectively.
Table 1 Analysis of factors influencing the decision-making behavior toward self-cultivation
Factor Model (l) Model (2) Model (3) Model (4) Model (5)
Male emigration -0.242*** -0.217*** -0.192*** -0.157** -0.154**
Young and middle-aged emigration -0.197*** -0.151** -0.141** -0.115* -0.105*
Educated people emigration 0.113** 0.084* 0.078 0.088* 0.082*
Surplus labor 0.170*** 0.153*** 0.146*** 0.147***
Non-farm income 0.140** 0.094 0.099*
Farmland fragmentation -0.259*** -0.262***
Distance to the nearest road -0.095*
_cons 0.989*** 0.936*** 0.801*** 0.833*** 0.848***
N 213 213 213 213 213

Note:*, **, *** indicate significance at the 10%, 5% and 1% levels, respectively.

4.3.3 Effects of population migration on farmland transfer behavior

The results (Table 2) showed that the factors that significantly influenced farmland transfer behavior were mainly emigration of males, young and middle-aged and educated people, and distance to the nearest road. Among these, emigration of male and educated people was positively significant at the 10% level, while emigration of young and middle-aged people was positively significant at the 1% level with farmland transfer, indicating that the loss of this main labor force was very important in the decision to transfer farmland, i.e., the more young adults leave, the greater the move to transfer farmland. Distance to the nearest road was positively significant with respect to farmland transfer at the 5% level, indicating that the access to transport was to some certain extent a catalyst for the decision to transfer.
Table 2 Analysis of the factors influencing the decision-making behavior toward farmland transfer
Factor Model (l) Model (2) Model (3) Model (4) Model (5)
Male emigration 0.050** 0.050** 0.050** 0.050** 0.038*
Young and middle-aged emigration 0.065*** 0.065*** 0.065*** 0.065*** 0.065***
Talent emigration 0.023* 0.022* 0.022* 0.022* 0.028**
Surplus labor -0.012 -0.012 -0.012 -0.008
Non-farm income 0.001 0.001 0.001
Farmland fragmentation -0.001 0.003
Distance to the nearest road 0.019**
_cons -0.013* -0.021 -0.002 -0.002 -0.019
N 213 213 213 213 213

Note: *, **, *** indicate significance at the 10%, 5% and 1% levels, respectively.

4.3.4 Effects of population migration on farmland abandonment behavior

The results (Table 3) showed that the factors influencing the farmland abandonment behav-ior were the emigration of males, young adults and educated people, surplus labor force and non-farm income. Among them, emigration of males, emigration of educated people and non-farm income all had a significant positive effect on the decision to abandon the farm at the 10% level, while the loss of young adults had a significant positive effect on the abandonment behavior at the 5% level. The proportion of surplus labor was significantly negatively related to farmland abandonment at the 1% level, indicating that the availability of surplus labor tends to prevent the decision to abandon farmland.
Table 3 Analysis of the factors influencing the decision-making behavior toward farmland abandonment
Factor Model (l) Model (2). Model (3) Model (4) Model (5)
Male emigration 0.170*** 0.159** 0.158** 0.146** 0.125*
Young and middle-aged emigration 0.140** 0.142** 0.131** 0.136** 0.140**
Talent emigration 0.140*** 0.118** 0.111** 0.097* 0.101*
Surplus labor -0.215*** -0.208*** -0.207*** 0.201***
Non-farm income 0.134* 0.131* 0.128*
Farmland fragmentation 0.034 0.053
Distance to the nearest road -0.057
_cons -0.051 0.150* 0.139* 0.120 0.143*
N 213 213 213 213 213

Note: *, **, *** indicate significance at the 10%, 5%, and 1% levels, respectively.

5 Discussion

5.1 The influence mechanism of farmland use behaviors from the perspective of population migration

Based on the results of the empirical analysis, different farmland use behaviors have been identified as being influenced by different mechanisms under the perspective of population migration, and this section will explore the influencing mechanisms that influence different individual farmland use behaviors.
(1) Mechanisms influencing self-cultivation decision-making behavior
With the expansion of male migration, the main labor force in the family engaged in agricultural production decreased, and the labor force engaged in agricultural production was dominated by women and the elderly, whose ability to farm was reduced by their daily activities of raising and caring for the family and their reduced strength/poor health, respectively; the loss of young adults, who are the main labor force, made the labor force inadequate for agricultural production. Due to the low income from agricultural production activities and the desire to improve their family's standard of living, young adults preferred to work in non-agricultural industries to maximize the benefits of their human capital, resulting in a reduced willingness to farm. When the population with high school or higher education was lost, they took with them a wide range of information channels and sources with them, as they know how to use advanced technology, and they have a greater ability to adopt and practice modern agricultural technology (Liu et al., 2022). Therefore, they will choose to use machinery to replace labor in agriculture, so that the proportion of agricultural income in the household income is increased (Yan et al., 2016). As a result, the emigration of educated people from the village promoted the decision to self-cultivation, which shows the characteristics of “leaving the village but not leaving the land”. The more surplus labor which was available in the village, the more farmers were able to allocate labor, with part of the labor force being engaged in agricultural production, leading to a higher degree of self-cultivation. Increased non-agricultural income provided more capital (Jansuwan and Zander, 2021), so that machinery could be increasingly used for agricultural production, which improved agricultural efficiency, reduced labor requirements and put the decision to self-cultivate on a firmer footing. The greater the fragmentation of farmland, the more material, human and financial resources are required for agricultural production, which increases the cost of agricultural production, and farmers will often decide against self-cultivation after weighing up the pros and cons. Furthermore, the closer the distance to the nearest highway, the better the farm's accessibility to markets and labor within commuting distance; if the agricultural benefits outweigh the transport costs, the chances of self- cultivation will be greatly increased, following the principle of maximizing benefits (Liu et al., 2022).
(2) Mechanisms influencing the decision-making behavior of farmland transfer
Increased male emigration means that the labor force engaged in agricultural production consists mainly of women and the elderly, encouraging families to transfer ownership of their farmland. Increased emigration of young and middle-aged people, who are always the main labor force in agricultural production, reduces the labor input per unit of agricultural production, which in turn reduces the agricultural production capacity and reduces the dependence on farmland, which in turn makes families more likely to transfer their farmland (Wang et al., 2018). The higher the emigration of educated people, the more of these people are lost; their dependence on agriculture is lower, and they tend to invest in high-income non-farm industries. In order to take advantage of national policies and thus increase their income sources, farmers tended to transfer their farmland, although the higher the number of surplus laborers, the weaker the willingness to transfer, in today's rural life.
When family labor force is abundant, farmers prefer to cultivate the land themselves rather than transfer it, which, in turn, indicates that farmland is still an important guarantee of a farmer's livelihood and there is a greater dependence on it. Higher non-agricultural income, reflecting a larger share of the population participating in higher-return non-agricultural industries (Zhu and Cai, 2016), leads to a decrease in the labor input in agricultural production, so the likelihood of transferring farmland increases accordingly. The greater the fragmentation of farmland, the lower the cost-effectiveness of farming the resulting parcel due to reduced access to irrigation and mechanized inputs, and fragmentation accelerates the rate of farmland transfer. The closer the farmland is to a main road, the more convenient is to transport, and the rural population is more inclined to go out to work rather than farm, and thus opting for farmland transfer.
(3) Mechanisms influencing the decision-making behavior of farmland abandonment
Men have always been the core labor force for household agricultural production activities. With the mass migration of men to work, the unit labor input for household agricultural production will decrease, resulting in less labor available for the original scale of production (Li et al., 2014), thus increasing the risk of farmland abandonment. Young adults are the main labor force, and their mass migration leaves a predominantly order labor force with limited capacity to work in agriculture. The more highly educated the emigrating population, the more this group tends to choose to go out to work in pursuit of high incomes and to maximize self-worth, thus increasing the rate of farmland abandonment. The more surplus labor is available, the less farmland will be abandoned (Guo et al., 2020). Farmers are rational economic people; with the spread of urbanization brought about by the improvement of human capital, faced with the imbalance between the high income of non-agricultural industries and the low income of agriculture, farmers are more willing to abandon their farmland (Li et al., 2014). The greater the degree of fragmentation, the more unfavorable the conditions for irrigation and the more difficult it would be to achieve mechanization (Mou et al., 2022); the higher the cost of labor input, and the higher the probability of farmland abandonment are also factors. The closer the farmland is to the nearest highway, the more accessible it is and the less likely that the farmer is to abandon their farmland. This reflects the fact that if the value of the land in terms of supporting the family is much higher than the costs of travel and living, the farmer will choose not to abandon the farmland.
(4) Comparative analysis of influence mechanisms on decision-making on farmland use
In the context of population migration, the effect of male emigration on farmland use was self-cultivation > transfer > abandonment. The effect of the migration of young adults on farmland use was transfer > abandonment > self-cultivation. The effect of the migration of highly educated people was transfer > self-cultivation > abandonment. These findings showed that all three types of farm behavior were influenced by human capital, although to different degrees. Specifically, the loss to the population of men was accompanied by the feminization and aging of the family farming labor force, with women and the elderly usually being physically weaker, less healthy and less experienced in farming, and this affects the degree of labor input for self-cultivation, which has a significant impact on the chances of its retention. Additionally, the transfer and abandonment of farmland were influenced by other production factors, and there was less transfer and abandonment in Longxi county overall. On the one hand, young adults as the primary labor force will affect the human factors involved in agricultural production. They will bring new ideas and insights from working outside the agricultural sector, making them more likely to move to urban areas. This move could ensure they have two sources of income and reduce transportation costs that occur when returning to farming. On the other hand, urbanization is developing, and as a result, human capital is increasing. This development highlights the advantages of young adults engaging in non-agricultural industries instead of agriculture. It prompts them to completely disengage from agricultural production, which leads to a rise in farmland abandonment. Therefore, the effect of young adult emigration on farmland abandonment behavior is also greater. With increased migration of highly educated individuals comes wider employment opportunities, greater social and economic integration with the city, and continual progress and broadening of ideas and insights. Such migration can increase the income of the family, whereas it can enjoy the convenience brought by various government policies, which makes them more inclined to farmland transfer. Therefore, the level of influence of migration on farmland transfer was higher.

5.2 Policy

As an important source of food, farmland seriously restricts the pace of agricultural transformation and rural revitalization, as well as food security and food safety. In order to prevent problems from getting worse, it is urgent to formulate corresponding measures to improve the poor use of farmland. Therefore, based on the research results, the following measures are proposed:
(1) Attracting farmers to return to their hometowns and encouraging new types of farmers (Li et al., 2022)
In order to suppress the phenomenon of bad land use, it is imperative to balance the rural labor force and agricultural production firstly, optimize the allocation of manpower structure, and achieve the optimal utilization of human resources. Secondly, governments concerned should foster and wholeheartedly support the entrepreneurism of returning farmers in their hometown (Guo et al., 2020). They should urge farmers to cultivate iconic plantation industries, whilst guaranteeing dependable support, thereby alleviating returnees’ psychological concerns. Later, providing agricultural technology training platforms is essential to boost farmers’ agricultural awareness (He et al., 2020), and cultivation standards to augment agricultural proceeds and augment enthusiasm in farming endeavors. Finally, since one of the purposes of labor migration is to seek more excellent living conditions, it becomes necessary to boost the building of rural basic public service systems, steadily enhance the standards of rural medical treatment and education, and offer more comprehensive security measures for farmers, so as to slow down the trend of labor migration.
(2) Moderate-scale operation, enhance farmers’ awareness of farmland transfer (Xie et al., 2023).
To begin with, scaling up operations can boost economies of scale (Wang et al., 2018), motivate farmers to pursue the idea of scaling up, enhance the intensive utilization of farmland, and thus increase the efficiency of farmland utilization. Nevertheless, it is essential to augment the investment made in different agricultural subsidies (Jansuwan and Zander, 2021). Secondly, since the implementation of agricultural subsidy policies often leads to a rise in land rent prices, the government should reduce its involvement in the development of scale management while increasing the proportion of scale management. Next, the relevant departments and policymakers should implement a transfer mode that complies with laws and regulations and generates real economic benefits based on the local economic development environment, while also establishing and enhancing the social security system under the farmland transfer system. Lastly, improve the publicity of farmland transfer, clarify to farmers the importance of such transfers, and subsequently raise their awareness towards it while standardizing their transfer behavior.
(3) The construction of agricultural infrastructure should be strengthened (Liang et al., 2020) and the mechanization level should be improved to comprehensively alleviate the abandonment of farmland.
First of all, due to the complex topography of Longxi county, irrigation facilities and transportation network construction should be vigorously carried out (Zhang et al., 2014) to ensure the smooth implementation of agricultural technology, and thus reduce the abandonment behavior of farmers caused by high grain growing costs. Secondly, the aging of the agricultural labor force is increasing due to the migration of young adults. It is essential to invest in agricultural technology and promote mechanization, develop low-cost small agricultural machinery suitable for hilly and mountainous areas to compensate for the shortage of elderly labor, and reduce the labor force required per unit of agricultural production. Finally, appropriate input of pesticides and fertilizers to promote the use of good varieties and prescriptions, improve the potential of farmland production and comprehensive grain production capacity, ensure grain output, promote food security.
(4) Strengthen the implementation of fallow rotation system to realize the sustainable use of farmland
Firstly, it is necessary to fully consider the supply and structure of agricultural products, and seek a reasonable crop rotation and fallow system according to local conditions, in order to ensure the balanced increase and efficiency of crops. Secondly, the relevant departments should popularize the system of fallow rotation and reinforce farmers’ awareness of the importance of fallow rotation. Lastly, to guarantee the successful implementation of fallow rotation, the protection of farmers’ basic rights and interests is crucial. A basic and practical subsidy distribution system should be implemented, and proper disbursement of funds issued by the Chinese government's financial department should reach the counties and townships and ultimately the farmers.

5.3 Innovation and shortcomings

Relative to the current literature, this paper presents a more comprehensive study of farmland use behavior under the perspective of population migration, which has been changing in recent years in response to increasing urbanization. This has an impact on various aspects of the countryside and agriculture and food security have always been the focus of global attention, so it is crucial to study the correlation between population migration and farmland use. This paper started from the village scale, which was more applicable and useful than the large-scale studies and was conducive to the subsequent agricultural development and the implementation of related agricultural policies. At this stage, there were different relationships between labor emigration and different farmland use behaviors. Although most of the previous studies focused on one behavior, the current paper used the data obtained to carry out an empirical analysis of three behaviors, namely self-cultivation, transfer and abandonment of farmland, and identifying differences in the impact mechanisms between different behaviors, thus adding considerably to the existing studies. However, there are still some shortcomings in the current research that need to be overcome, mainly in that, on the one hand, farmland use behavior is not only influenced by the population but also by the level of mechanization and related agricultural policies; due to the limited access to data, this aspect has not been studied, so a more comprehensive analysis needs to be carried out. Furthermore, the dynamic characteristics of the populations also affected the behavior of farmland use, which was only studied from a cross-sectional perspective in the current paper, so a more comprehensive study from the spatial and temporal perspective of the population is needed in the future. Finally, this paper mainly studies from the perspective of population migration. In reality, as the country continues to promote the process of agricultural and rural modernization and rural revitalization, new agricultural business entities continue to emerge in rural areas and become an important force in this process. The new management subject can not only increase the income of farmers, but also realize the full utilization of land resources, thus promoting the transfer of farmland and reducing the abandonment of farmland. Therefore, its role should be taken into consideration in the future research.

6 Conclusions

This paper constructed a decision model of farmland use behavior, revealed the current situation of differences among three types of farmland use behavior from the perspective of population migration and the corresponding influence mechanisms, and conducted an empirical study using survey data from 213 villages in Longxi county, and described the main results obtained.
(1) In the population of Longxi county, the migration of males, young and middle-aged and educated people showed significant disparities in quantity and distribution, in which the degree of male and young and middle-aged migration is mainly higher and lower value, whereas the degree of educated people migration was more evenly distributed. Simultaneously, male migration spatially presented a distribution of high in all directions and low in the middle, with young and middle-aged migration highest at the north and south ends and low in the central areas, with migration of educated people being highest around the G30 national road at high-value areas in the west and at lower-value areas in the east.
(2) The primary farmland use behavior in Longxi county was self-cultivation, followed by farmland abandonment and finally transfer. Among the 213 villages studied, 179 villages (84.04%) had an area of self-cultivation higher than 60%, 164 villages (77.00%) had an area of farmland transfer lower than 5%, and 179 villages (84.04%) had an area of farmland abandonment that was less than 30%.
(3) From the perspective of population migration, there were significant differences in the factors influencing different farmland use behaviors, among which the extent of emigration of males, young adults and educated people were the common factors influencing the three types of farmland use behavior. Emigration of either males or young adults significantly inhibited the self-cultivation behavior. Conversely, the emigration of educated people significantly promoted self-cultivation behavior. The emigration of all three labor force constituencies significantly promoted the farmland transfer and abandonment behaviors. In addition, the factors influencing self-cultivation (homesteading) behavior included the proportion of surplus labor, non-farm income, fragmentation of cultivated land, and distance from the nearest road, whereas the factors influencing farmland transfer behavior included the distance from the nearest main road, and the factors influencing farmland abandonment behavior included the proportion of surplus labor and non-farm income.
Alcantara C, Kuemmerle T, Baumann M et al., 2013. Mapping the extent of abandoned farmland in Central and Eastern Europe using MODIS time series satellite data. Econstor Open Access Articles, 8(3): 1345-1346.

Baumann M, Kuemmerle T, Elbakidze M et al., 2011. Patterns and drivers of post-socialist farmland abandonment in western Ukraine. Land Use Policy, 28(3): 552-562.


Benayas J M R, Martins A, Nicolau J M et al., 2007. Abandonment of agricultural land: An overview of drivers and consequences. CAB Reviews: Perspectives in Agriculture, Veterinary Science, Nutrition and Natural Resources, 57(2): 1-12.


Benjamin D, Brandt L, 2002. Property rights, labor markets, and efficiency in a transition economy: The case of rural China. Canadian Journal of Economics, 35(4): 689-716.


Deininger K, Jin S, 2008. Land sales and rental markets in transition: Evidence from rural Vietnam. Oxford Bulletin of Economics and Statistics, 70(1): 67-101.


Deininger K, Jin S, Nagarajan H K, 2008. Efficiency and equity impacts of rural land rental restrictions: Evidence from India. European Economic Review, 52(5): 892-918.


Dong H, Zhang Y N, Sun Y Q et al., 2022. To keep or not to keep the farmland? Incentives and barriers to farmers’ decisions in urbanizing China. Habitat International, 130: 102693.


Guo B B, Fang Y L, Zhou Y K, 2020. Factors influencing farmland abandonment and spatial differentiation at the farm household scale. Resources Science, 42(4): 696-709. (in Chinese)

He Y F, Xie H L, Peng C Z, 2020. Analyzing the behavioral mechanism of farmland abandonment in the hilly mountainous areas in China from the perspective of farming household diversity. Land Use Policy, 99: 104826.


Huy H T, Lyne M, Ratna N et al., 2016. Drivers of transaction costs affecting participation in the rental market for cropland in Vietnam. Australian Journal of Agricultural and Resource Economics, 60(3): 476-492.


Jansuwan P, Zander K K, 2021. What to do with the farmland? Coping with ageing in rural Thailand. Journal of Rural Study, 81: 37-46.


Jin S Q, Jayne T S, 2013. Land rental markets in Kenya: Implications for efficiency, equity, household income, and poverty. Land Economics, 89(2): 246-271.


Kozak J, 2003. Forest cover change in the western Carpathians in the past 180 years: A case study in the Orawa Region in Poland. Mountain Research and Development, 23(4): 369-375.


Li F, Ma S, Liu X L, 2023. Changing multi-scale spatiotemporal patterns in food security risk in China. Journal of Cleaner Production, 384: 135618.


Li L F, Khan S U, Guo C H et al., 2022. Non-agricultural labor transfer, factor allocation and farmland yield: Evidence from the part-time peasants in Loess Plateau region of Northwest China. Land Use Policy, 120: 106289.


Li Y L, Zhou T, Jiang G H et al., 2021. Spatial pattern and mechanisms of farmland abandonment in agricultural and pastoral areas of Qingzang Plateau. Geography and Sustainability, 2(3): 139-150.


Li Z H, Yan J Z, Hua X B et al., 2014. A study on different types of farm abandonment and their influencing factors: A case study of 12 typical villages in Chongqing. Geographical Research, 33(4): 721-734. (in Chinese)

Liang X Y, Li Y B, Zhou Y L, 2020. Study on the abandonment of sloping farmland in Fengjie county, Three Gorges Reservoir area, a mountainous area in China. Land Use Policy, 97: 104760.


Liao L W, Long H L, Ma E P, 2021. Changes in rural labor factors and transformation of arable land use. Economic Geography, 41(2): 148-155. (in Chinese)

Liu C H, Nie W J, Zhao X T et al., 2022. Regional differences in the disposition of contracted land of rural migrant workers and the influencing factors: Based on social integration perspective. Journal of Natural Resources, 37(2): 424-439. (in Chinese)


Liu Y S, 2018. Urban-rural integration and rural revitalization in the new era of China. Acta Geographica Sinica, 73(4): 637-650. (in Chinese)

MacDonald D, Crabtree J R, Wiesinger G et al., 2000. Agricultural abandonment in mountain areas of Europe: Environmental consequences and policy response. Journal of Environmental Management, 59(1): 47-69.


Mou Y, Zhao Y L, Li X B et al., 2022. The influence of plot quality characteristics on terrace abandonment in southwestern mountainous areas: An example from Baidu village, Jianhe county, Guizhou. Geographical Research, 41(3): 903-916. (in Chinese)

National Bureau of Statistics of China NBSC, 2020. 2020 migrant workers monitoring survey.

Otsuka K, 2007. Efficiency and equity effects of land markets. Handbook of Agricultural Economics, 3: 2671- 2703.

Ramankutty N, Foley J A, Olejniczak N J, 2002. People on the land: Changes in global population and croplands during the 20th century. AMBIO: A Journal of the Human Environment, 31(3): 251-257.


Rogers S, Wilmsen B, Han X et al., 2021. Scaling up agriculture? The dynamics of land transfer in inland China. World Development, 146(4): 105563.


Shao J A, Zhang S C, Li X B, 2015. The role of land transfer in mountainous areas in alleviating arable land abandonment. Acta Geographica Sinica, 70(4): 636-649. (in Chinese)

Song S X, Liang X Y, Chen H et al., 2018. Simulation study of arable land abandonment based on multi-intelligence body and land conversion model: Taking Mili county of Shaanxi province as an example. Journal of Natural Resources, 33(3): 515-525. (in Chinese)

Wang G G, Liu Y S, Chen Y F, 2014. Analysis of the trend and driving forces of intensive arable land use in China. Acta Geographica Sinica, 69(7): 907-915. (in Chinese)

Wang J Y, Cao Y, Fang X Q et al., 2022. Does land tenure fragmentation aggravate farmland abandonment? Evidence from big survey data in rural China. Journal of Rural Study, 91: 126-135.


Wang J Y, Li X B, Xin L J, 2018. Study on the spatial and temporal evolution characteristics and influencing factors of land transfer in China. Journal of Natural Resources, 33(12): 2067-2083. (in Chinese)


Wang Y B, Zhao P F, 2022. Non-farm employment, agricultural productive services and farmland transfer: A survey data based on rice farmers in Xiangwansu. China Agricultural Resources and Zoning, 43(2): 113-121. (in Chinese)

Wang Y H, Li X B, Xin L J et al., 2018. Regional differences in land transfer in China and its influencing factors: Based on data from rural fixed observation sites from 2003-2013. Acta Geographica Sinica, 73(3): 487-502. (in Chinese)

Wang Y H, Li X B, Xin L J et al., 2020. Farmland marginalization and its drivers in mountainous areas of China. Science of The Total Environment, 719: 135132.


Wu S S, Ma L B, Tao T M et al., 2022. Structure and governance model of rural social space quality: A case study of Longxi county in the loess hilly area of China. Journal of Geographical Sciences, 32(7): 1297-1320.


Xie H L, Huang Y Q, 2022. The influence of non-farm employment and land transfer on the abandonment behavior of farmland by farming households: A case study in the mountainous area of Min-Gan-Xiang. Journal of Natural Resources, 37(2): 408-423. (in Chinese)


Xie H L, Wu Q, Li X B, 2023. Impact of labor transfer differences on terraced fields abandonment: Evidence from micro-survey of farmers in the mountainous areas of Hunan, Fujian and Jiangxi. Journal of Geographical Sciences, 33(8): 1702-1724.


Xu D D, Deng X, Guo S L et al., 2018. Labor migration and farmland abandonment in rural China: Empirical results and policy implications. Journal of Environmental Management, 232(15): 738-750.


Yan J Z, Yang Z Y, Li Z H et al., 2016. Drivers of cropland abandonment in mountainous areas: A household decision model on farming scale in Southwest China. Land Use Policy, 57: 459-469.


Zhang Y, Li X B, Song W et al., 2014. Different scale effects of agricultural labor on arable land abandonment under agricultural land transfer in Wulong county, Chongqing. Progress in Geography, 33(4): 552-560. (in Chinese)

Zhao L J, Kang X H, Shi J H, 2021. The impact of farmland transfer on farmers’ household poverty vulnerability and its regional variation analysis. Journal of Natural Resources, 36(12): 3099-3113. (in Chinese)


Zhu L L, Cai Y Y, 2016. Influence of farm household livelihood endowment on agricultural land transfer: An example of different types of functional areas in Hubei province. Journal of Natural Resources, 31(9): 1526-1539. (in Chinese)