Introduction

Changes in the pasture management system set the historical background for the occurrence of pasture transfer behaviour of pastoralists. The pasture management system has undergone three significant adjustments from 1949 to the present, especially the "grass-livestock double contract" system implemented in the mid to late 1980s, which formed a livestock management system with the herders’ family contract as the core. "Grass and livestock" are divided into households according to population, which inspires herdsmen’s enthusiasm for production. At the same time, herders face new development dilemmas, such as the degradation of grassland carrying capacity and ecological environment due to fineness and overload1,2. Some pastoralists are gradually choosing to move out of grasslands and into towns for alternative livelihoods for subsistence development. Over the past decade, the Chinese government has introduced and implemented policies such as the "Western Development Strategy," the "Belt and Road," as well as the "Restored to forests or grasslands" and the "Grassland Ecosystem Compensation and Incentive Mechanism" to promote economic development in the Northwest of China, leading to increasingly frequent migration of herdsmen between urban and rural areas. Scholars have found through quantitative studies of pasture flow in Qinghai3, Xinjiang4, and Inner Mongolia that pasture flow exists to varying degrees in each region. Among them, the incidence of streaming in Qinghai Province was 2.35% in 20025, which played a positive role in developing grassland animal husbandry. Li Xiandong et al.4 found that the incidence of grassland streaming in sampled herder households in Xinjiang was 27.2% in 2017–2018, and the participation rate continued to rise. The incidence of streaming in Ulanchab, Ordos, and Xilinguole League of Inner Mongolia in 2021, as well as Hulun Buir in 2023, reaches 51.68%6. The incidence of pasture turnover has been increasing, and it is predicted that pasture turnover among herders will still further increase, so it is of great significance to study the driving factors affecting pasture turnover among herders.

Pastoralist pasture transfer behaviour varies according to the pastoralist’s sentiments and social differentiation. The author’s research found that more than 60% of herders who left the grassland to work in the city did not have the will to turn out of the pasture. In comparison, the proportion of herders who stayed in the meadow and operated livestock willing to turn out of the range was about 30%. The unique emotions of farmers and herders towards a place and the pastoral division within the family are essential factors in the formation of place attachment. The traditional economic model of nomadic herders relies mainly on animal husbandry, and the migration of populations due to climatic shifts often promotes cultural exchanges and ethnic integration, forming a special bond with the pasture. However, the transformation of the livestock economy has prompted herders to engage in non-livestock labour, indirectly affecting the degree of pasture use. As7 and Tian et al.8 have shown, farmers’ perceptions and income and labour are the main reasons influencing the transfer of agricultural land. Similarly, most ranchers in the United States enjoy the ecological landscape of their pastures, and even if they no longer rely on the income from their ranching operations, they still possess a strong desire to maintain their ranch operations, and as much as 45% of grass pastures are brought to the market for sale every decade2,9. Enhancing herders’ willingness to transfer pastureland will alleviate the impact of pasture fragmentation on ecological degradation and the decline in grassland biodiversity caused by overloading and overgrazing. Improving the pasture use system and increasing pasture transfer among herders is one important way to achieve effective resource allocation and promote the development of modern animal husbandry.

The herders’ attachment to the grassland and the mobile mode of production makes the herders have strong emotions towards the grassland. However, the economic and social policy reforms also give the herders’ families different degrees of social differentiation, directly or indirectly affecting how the herders’ grassland is utilized. Although there are many scholars on the transfer of agricultural land factors affecting research, relatively ignore the herders themselves for the grassland emotion, prompting herders, even if the livestock economy is not high, not have the time and energy to manage the pasture not willing to turn out of the pasture. The differentiation of family social status leads to the occurrence of herders’ part-time behaviour, and the potential segmentation and fine-grained fragmentation pattern reflects different degrees of ownership fragmentation of the pasture, which seriously threatens the ecological environment2. Therefore, it is necessary to address the impact of ethnic sentiment and social differentiation on grassland turn-out. Given this, based on controlling the price of pasture turnover, the research in pastoral areas of Inner Mongolia and Xinjiang is used as an example to analyze the willingness of pasture turnover of border ethnic minority herders and the factors affecting it, and then provide policy references to promote pasture turnover, increase the income of herders and promote the sustainable development of pastoral areas.

Literature review

Land system reform has been a critical area of long-term attention for the Chinese government and scholars, and rich results have been achieved in rural land transfer in particular10, focusing mainly on transfer willingness and influencing factors10,11,12, agricultural productivity and resource allocation13,14,15, household income and welfare improvement16,17,18, and labour mobility19,20. Since pasture belongs to a kind of agricultural land and pasture flow is a unique form of land flow, the research idea of pasture flow is similar to land flow. However, few scholars have focused their research perspectives on the effects of ethnic sentiment and social differentiation on pastureland transfer among herders.

Ethnic sentiment is the spiritual pillar pastoralists rely on for survival

Fei21 refers to the farmers’ deep sense of belonging to the countryside, where they have lived for generations and the land they depend on for survival as the vernacular complex. Gentle12 argues that pasture turnover is a traditional form of pasture utilization associated with culture and ethnicity. However, some studies show that pastoralists’ unique attachment to grassland is also an essential factor in a low rate of pasture turnover22,23. Strong ethnic sentiments and social relations between family and neighbours are the main reasons for pastoralists to maintain their herding, and farmers’ attachment to their working and living landscapes leads to a reluctance to transfer their land even when they are "off the land and out of the country"24,25. Field research and interviews found that herders, like farmers, graze their livestock in the grasslands, flourish and have a long history and culture, and form a social culture very different from that of farmers, which is also one of the main factors that herders are reluctant to transfer their pastures26,27. The research on the land leasing market in Vietnam, India, and Ethiopia from the perspective of gender, religious belief, cultural customs, and family decision-making power also basically demonstrates that the religious culture of various ethnic groups has different degrees of influence on land leasing, land use, and ecological protection28,29,30. Due to the differences in the emphasis of scholars and research areas, the factors that influence the circulation of grassland show different characteristics. It is undeniable that scholars affirm the positive role of pastureland transfer in realizing resource matching, developing middle-scale management and protecting grassland ecology, and helping to promote the evolution of livestock production division of labour, enriching the livelihood options of herders, and improving farmers’ income and quality of life31,32. Still, the influence of national feelings on the circulation of grassland is relatively ignored.

Social differentiation is an endogenous motivation for pastoralists to explore livelihood development

With the settlement of herders and the government’s attention to ecological protection, significant changes have occurred in the economy and society of pastoral areas. Significantly, herders working in cities, their children attending school in cities, and choosing jobs are becoming increasingly common. The change of livelihood mode promotes the social differentiation of herders and significantly affects the willingness of herders to switch out of grassland33. Regarding the definition of social differentiation of pastoralists, this paper mainly relies on the theory of social differentiation and inductively deduces the definition of social differentiation of pastoralists from the concepts of differentiation, social differentiation, social stratification, and social mobility. That is the process by which pastoralists develop from a particular social position in the structure of the social system, which initially assumed multiple functions, to various social positions that assume a single function. Leads to a greater variety of options for pastoralists, and they may turn to agricultural production, tourism, handicraft production, etc., to increase their sources of income. However, those herders who rely solely on grazing for their income often opt for large-scale grazing to maximize their profits34,35. However, in the absence of sufficient external sources of income, even if grazing scales up, grass-pasture enterprises face significant livelihood challenges if profits fall below zero. Under heavy economic pressure, ranchers may need to scale down their operations or turn out of the pasture. When the number of herders’ family part-timers is high, the pastureland is in a state of idleness or inefficient utilization, which will prompt herders to transfer the pastureland to capable business entities for large-scale operation36,37. On the contrary, when the number of family herders is high, animal husbandry is the primary source of family livelihoods, and herders not only do not transfer out of pasture but may also have pasture transferred in38. The essence of existing research on herder family income structure is based on the perspective of social differentiation, which is reflected in the fact that herder income and labour force produce different degrees of differentiation. However, existing studies lack a unified exploration of the two, such as Yi et al.39 only discussed the status of the number of herders’ household labour force, age structure, and production division, and Liu et al.40 only used the number of the available household labour force as a control variable to analyze the impact of social capital on pasture rental prices, all of which are explored in a single perspective and are not linked to the flow of pastureland.

Other influencing factors

Other scholars have argued that ecological constraints41, social security4, and flow revenue expectations42 have a significant impact on pastoralist pasture flow. Further research shows that the factors affecting pastureland transfer include transfer policy, income structure, labour structure, age, gender, and years of education of the household head4,27,43,44. The ageing problem is also a constraint on transfer behaviour. Based on controlling gender, age, and natural endowment, Xue et al.43 believed that circulation policies and income structure significantly impacted the circulation of herders’ grassland. In addition to policy and income structure, the structure of the labour force is also one of the critical factors that affect the production decisions of herders. Ranchers in the western United States tend to be seriously ageing. Huntsinger et al.45 found that the average age of California ranchers is 59. Peterson and Coppock46 found that 37% of the respondents were 66 years old or above in the survey of Utah livestock breeders. The proportion of young operators is declining, and the ageing phenomenon is increasing47. The older the rancher, the deeper the emotion for the pasture, manifesting itself in a stronger ethnic sentiment. Ranchers would rather risk their business or their lives than transfer their pasture. They often hand over daily activities to their children and maintain the ranges that their families may have for generations48. Ranchers pass their land to their offspring, and the degree of social differentiation gradually becomes apparent. The offspring’s sentiment for the land is lower than that of the ranchers. Most of them need to put more energy into running the ranch and tend to choose multiple livelihoods, which in turn affects the willingness to transfer pastures.

Literature synthesis

Through the literature review, it is found that the factors driving grassland transfer of herders are not only affected by individual and family characteristics but also by external policies and interest-driven factors. Although most studies also mention the influence of herders’ custom differentiation and labour structure on grassland circulation, they need to pay more attention to it. Based on the investigation of 14 counties, such as Inner Mongolia and Xinjiang, this paper finds that pastoral areas are inhabited mainly by ethnic minorities, such as Mongolian, Kazak, Tajike, and other famous ethnic groups, which have a long nomadic history, and strong ethnic feelings significantly affect herders’ willingness to turn out of grassland. At the same time, the change in economic and social structure in pastoral areas leads to the increasingly significant social differentiation of herders. It affects the choice of livelihood strategies of herders’ families. Given this, the paper analyzes herders’ willingness to transfer grassland from ethnic feelings and social differentiation.

Theoretical analysis and research hypothesis

Ethnic sentiment and pastoralists’ willingness to turn out of the pasture

Some scholars have analyzed the impact of farmers’ participation in farmland circulation from the perspective of their rural attachment, but more research is needed on the circulation of grassland among herders. Research conducted in pastoral areas found that unique emotions or cultural customs formed based on nomadic history significantly impact herders’ livestock production. Nomadic life is not only a mode of production but also a way of life and an emotional bond, with the same culture and belief system leading to the retention of similar ways of living over time. Chi et al.49 pointed out that national cultural identity is one of the important drivers of Land Transitions, which can enhance herders’ sense of national cultural identity and pride. It can also deepen the understanding and expression of ethnic sentiment and promote the emotions and connections between herders. Nomadic life, with its exceptional natural environment and lifestyle, requires the formation of close social ties and collaboration among herders50. During nomadic pastoralism, pastoralists are concerned with the ecological environment of the grasslands, the economic benefits of their livestock, and the sense of community among pastoralists, as well as the emotional connection and interdependence with their grassland livestock, families, and neighbors51. This network, linked by the grassland, provides emotional comfort and livelihood security for herders, who support each other in times of difficulty and work together to cope with natural disasters and economic pressures38. Mutual assistance among herders is not only emotional support but is also reflected in practical resource sharing and cooperation, resulting in a highly interdependent way of life. Nomadic sentiments make herders emotionally dependent on grassland life, which is difficult to give up easily. In contrast, mutual aid sentiments make herders feel a sense of loss of collective security when they turn out of the pasture by reinforcing the social network among herders. Combining the two creates a complex interaction of emotion with the economy and society.

Even against the rapid development of modern society and the significant changes brought about by technological innovations, pastoral areas still maintain a traditional nomadic life. Herders move their livestock in search of fresh pastures and water sources. This method effectively promotes ecological balance, protects grassland ecology, and establishes a particular emotion between herders and pastures. During herders’ migrations, due to the large size of the grassland and the scarcity of people, herders have a close relationship of mutual assistance, sharing resources while helping each other solve difficulties. The herders’ unique emotion towards the pasture makes them more cautious when they carry out the transfer out of the pasture. Explicitly speaking, herders have been influenced by nomadic culture for a long time, forming relatively consistent social identification, values, or psychological sense of belonging within the tribes, which is the national feeling of herders. Traditional grassland animal husbandry requires nomadic herding to complete livestock production and transhumance within a relatively fixed time frame. The older the herders are, the deeper their nomadic impression is, and the more they expect that the traditional nomadic culture can be inherited and developed to ensure the natural restoration of grassland ecology. However, if the grassland encounters drought, snow disaster, or rat damage, the herders in the tribe have the tradition of spontaneously helping each other to overcome difficulties. In daily animal husbandry production, herders will also cooperate to complete production activities such as grass cutting, transfer, and wool shearing, which can significantly increase the disaster resistance ability of herders. We define the former as nomadic sentiment and the latter as mutual aid sentiment and then put forward research hypothesis 1: nomadic sentiment and mutual aid sentiment significantly negatively affected the willingness of herders to turn out of the pasture.

Social differentiation and pastoralists’ willingness to turn out of the pasture

The rural social structure is also evolving with China’s rapid economy and societal development. China has been strengthening the reform of rural grassroots organizations and the innovative development of social management based on the current rural social governance structure stage. This has been highlighted in the social differentiation of farmers and herders and has affected the land transfer market. The Tibetan Plateau in western China is undergoing a process of social differentiation52, where improved infrastructure and the development of towns and cities have altered economic incentives, with herders often choosing to transfer pastures out of their grazing areas and an increasing number of herders living part-time out of their grazing areas. Similarly, the process of social differentiation has accelerated in Isiolo, northern Kenya, with significant out-migration and land conversion, often linked to agricultural activities in small towns, and in Sardinia, Italy, where the traditional pastoral system has undergone necessary restructuring, with ecological changes of out-migration and land conversion due to changes in grazing patterns52. Peasant differentiation is regarded as the result of the natural evolution of the division of production in the process of land system innovation and socio-economic development, as well as the key to influencing the social differentiation of peasants in each region53. Nie and Zhong54 explained the influence of differentiation degree on farmers’ land transfer behaviour from horizontal and vertical differentiation. Social differentiation in Nepal’s mountainous regions facilitates the diversification of livelihoods in areas other than agriculture and pastoralism, with increased land turnover leading to a gradual shift from pastoral to non-pastoral livelihoods55.

Research in pastoral areas has shown that, along with the effective implementation of the pastoral settlement project, the economic structure of pastoral areas has gradually been upgraded and transformed, and the differentiation of pastoralists has become increasingly evident, the most noticeable feature being the changes in the structure of pastoralist families’ income from herding and the structure of the labour force’s allocation. The change in revenue structure means a change in leading business. Constrained by herders’ household endowments, the amount of energy devoted to livestock production is bound to wane if herders focus mainly on running their stores. Similarly, if the main household labour force is allocated to urban non-pastoral areas, it will also affect the operation of grassland animal husbandry. However, the research on the influence of social differentiation on herders’ grassland circulation is relatively insufficient. Drawing on Nie and Zhong’s definition of social differentiation of farmers (2014), the proportion of animal husbandry income and the proportion of animal husbandry labour force were used to represent the differentiation of herders. Therefore, research hypothesis 2 is that the proportion of animal husbandry income and animal husbandry labour force significantly negatively impact herders’ willingness to turn out of grassland.

The serious challenge of pasture degradation has prompted the state to introduce relevant policies to rectify the situation. At the same time, the transfer of pastureland by herders is an effective pasture protection measure1. However, ethnic sentiment and social differentiation in herders’ groups have different degrees of influence on herders’ willingness to turn out of the pasture. Based on the above analyses, the theoretical framework of this paper is constructed (see Fig. 1).

Fig.1
figure 1

A theoretical framework map of ethnic sentiment, social differentiation and pastoralists’ willingness to turn out of pastureland.

Full size image

Research scheme design

Data source and sample structure

The data comes from the investigation conducted by the research group in Inner Mongolia and Xinjiang Animal Husbandry Banner County in October 2018 and August 2022. The questionnaire design framework, survey content, and respondents of the two studies were similar, and both used stratified and random sampling to ensure the representativeness and consistency of the samples. The herders in the surveyed area are mainly from the Mongolian and Kazakh ethnic groups, having long histories of nomadic life and sharing similar traditional cultures. The specific sampling process is as follows: first, according to the geographic location of Inner Mongolia, one league was randomly selected from the east, central, and west; second, according to the pasture area and economic development of each league, each league was divided into several echelons, and a flag was randomly selected from each echelon as a sample flag; again, the soum of each sample flag was divided into three groups of high, medium and low according to the per capita grassland area of the soum as well as the status of livestock production, and a soum was randomly selected from each group as a sample soum. From each group, one Sumu was randomly chosen as the sample Sumu; after that, from each sample Sumu, one Gacha with a higher per capita grassland area than that of the Sumu and one Gacha with a lower per capita grassland area than that of the Sumu were randomly selected as the sample Gacha; finally, according to the size of the Gacha, 15–20 households were randomly selected as the sample herding households. Two hundred thirty herding households in 12 Gachas in 6 banners of 3 leagues were chosen for the survey in Inner Mongolia, and 21 invalid samples were excluded from obtaining 209 valid questionnaires.

The sample selection in Xinjiang was conducted within the latitude range (39°N—48°N) of the sample area in Inner Mongolia. Considering the large number of counties and cities in the region and the relatively concentrated distribution of pastureland, the group’s sample selection for Xinjiang was roughly divided into three stages. In the first stage, five prefectures in Xinjiang, namely Ili Kazakh Autonomous Prefecture, Tacheng, Altay, Changji Hui Autonomous Prefecture, and Bayinguoleng Mongol Autonomous Prefecture, were selected as the sample area, taking into account the differences in the number of livestock and the level of economic development; in the second stage, 1–2 counties were selected according to the area of pastureland and household size of each prefectural prefecture; in the third stage, 1–2 villages were randomly selected from each of the counties selected, and 15–20 families were selected for the study in the research villages based on the number of households. In the third stage, 1–2 villages were randomly selected in each of the chosen counties, and 15–20 households were selected in the research villages for the research. Four hundred eighty questionnaires were distributed in Inner Mongolia and Xinjiang, and 437 valid questionnaires were obtained, with an efficiency rate of 91.04%. The questionnaires covered individual characteristics of herders and basic family information, livestock production and operation status, and pasture flow status.

According to the sample sampling design, the proportions of valid samples obtained from Hulun Buir League, Xilin Gol League and Alxa League in Inner Mongolia were 17.4%, 16.9%, and 13.5% of the total valid samples, respectively; the proportions of valid samples from all states in Ili, Tacheng, and Altay in Xinjiang were 13.0%, 12.8%, and 17.2% of the total valid samples, respectively. The proportions of Changji Hui Autonomous Prefecture and Bayingoleng Mongolian Autonomous Prefecture were both 4.6%. In terms of the individual characteristics of the sample, 77% of the model were male, and 23% were female; the age was mainly distributed between 20–40 years old; the proportion of the interviewed sample was 92% for local ethnic minorities and 8% for herders of other (mainly Han) ethnic groups; the education level was concentrated at junior high school level. Regarding the characteristics of the sample households, those with family members holding village and above village public offices accounted for 12% of the total sample; the household water supply and electricity supply were in good condition. Regarding the characteristics of the sample households, those with family members holding village and above village public offices accounted for 12% of the total sample; the household water supply and electricity supply were in good condition. Regarding household livestock production characteristics, the average distribution of livestock assets converted into sheep units was between 20,100 and 50,000; household livestock rearing was converted into approximately 350 sheep units; herders’ expectations of livestock business prospects were relatively optimistic (see Table 1 for details).

Table 1 Variable definition and descriptive statistics.
Full size table

Variable selection and descriptive statistics

Dependent variable

Willingness to turn out of the pasture: Since the sample area involves 14 banners and counties in 8 leagues and cities in Inner Mongolia and Xinjiang autonomous regions, and the level of economic development varies greatly, to exclude the influence of pasture transfer price, the questionnaire emphasizes "Are you willing to transfer out of the pasture soon or in the future under the current market price of pasture transfer?" The answer to “willing” was recorded as "1", and the response to “unwilling” was recorded as "0". The sample statistics show that nearly 31% of herders want to switch out of the pasture.

Core independent variables

Identification of ethnic sentiment: As there are more factors to be measured in ethnic sentiment, factor analysis can simplify multiple observed variables into a few factors and reduce the complexity of the data set56. Therefore, this paper tries to adopt the factor analysis method, referring to the research results of scholars12,57 and combining with the actual situation of the pastoral area to set seven indicators to extract the common factors to characterize the herders’ ethnic sentiment variables. Generally speaking, through long-term nomadic life, herders form unique national history and cultural traditions, which become a spiritual support and emotional bond12, thus influencing their children to engage in livestock production activities58. At the same time, due to the low population density of the grasslands and their dispersed settlements, herders often find it challenging to get quick assistance from the outside world when they encounter difficulties. Therefore, mutual assistance among herders is even more critical59,60.

Factor analysis was performed on the above seven indicators with the help of SPSS software. The common factors were extracted by principal component analysis, and the KMO value of all measured question items was 0.676, and the Bartlett spherical test passed the 1% significance level test. Table 2 shows the total variance scale of analysis and interpretation of ethnic sentiment factors. The two significant variables with eigenvalues greater than one were extracted to explain 35.33% and 28.16% of the variance of the variables, respectively, reaching a total of 63.49%.

Table 2 Total variance scale of analysis and interpretation of ethnic sentiment factors.
Full size table

After their maximum variance pivoting, the two significant variables’ relative positions and cumulative variance contributions combined remained unchanged. The factors with components more effective than 0.7 were extracted from the rotated component matrix, and the years of grazing, whether they were nomadic, and whether they wanted their children to continue grazing constituted factor I: nomadic sentiment; the herders’ neighbour relationship and willingness to help each other formed factor II: mutual aid sentiment (see Table 3).

Table 3 Identification Scale of ethnic sentiment.
Full size table

Identification of Social differentiation: Drawing on Nie and Zhong54 on farmer differentiation, the social differentiation of pastoralists is characterized by the proportion of animal husbandry income and the proportion of animal husbandry labour force. Herders’ income was divided according to the proportion of herders’ livestock income to total revenue. Herders with a balance of 0–20% were defined as non-herders = 1, herders with a ratio of 21%-79% were described as part-time herders = 2, and herders with a balance of 80% or more were defined as pure herders = 3. Herdsman’s labour force differentiation, the proportion of labour force engaged in animal husbandry in herders’ families, is measured by measured values.

Control variables

The demographic characteristics of the sample individuals, gender, age, ethnicity, and education level; the household characteristics, whether the household members were seriously ill, whether the household members had a village and above village public officials, and the household water and electricity supply status; and the livestock production characteristics, such variables as livestock fixed assets, livestock and sheep units raised, and expectations of livestock business prospects were selected as control variables.

Model selection and setting

This paper examines the influence of ethnic sentiment and social differentiation on herders’ willingness to switch out of the pasture. The explanatory variable is "herders’ willingness to participate in switching out of pasture", which is a dichotomous "0, 1" variable. Probit and Logit models are more commonly used for binary selection models. Since the cumulative distribution function of the logistic distribution has an analytic expression, the Logit model is more appropriate. "Willingness to switch out of pasture" as the explanatory variable was estimated by constructing a Binary Logit model based on the herders’ micro-level (see Formula 1).

$$L = Ln\left( {\frac{p}{1 - p}} \right) = \beta_{0} + \mathop \sum \limits_{i = 1}^{m} \beta_{i} X_{i} + \mathop \sum \limits_{i = 1}^{m} \gamma_{i} Z_{i} + \mu$$
(1)

p represents the probability that the herders participated in the conversion, (1 − p) represents the probability that they did not participate in the transformation, and \(\frac{\text{p}}{1-\text{p}}\) is the odds. \({\text{X}}_{\text{i}}\) and \({\text{Z}}_{\text{i}}\) denote the core independent variables (nomadic feelings, mutual aid feelings, proportion of animal husbandry income and balance of animal husbandry labour force) and control variables (gender, age, ethnicity, education level, etc.) , respectively. The partial regression coefficient (\({\upbeta }_{\text{i}},{\upgamma }_{\text{i}}\)) of the Logit model is the change of log occurrence ratio (\(\text{Ln}\frac{\text{P}}{1-\text{P}}\)) caused by one-unit change of the corresponding independent variable and control variable. For ease of explanation, take the logarithm of e on both sides of formula (1) to obtain formula (2). \({\text{e}}^{{\upbeta }_{\text{i}}}\) is the odds ratio, which indicates the multiple of the odds ratio when the independent variable changes by 1 unit.

$$Odds = \frac{p}{1 - p} = \exp \left( {\beta_{0} + \mathop \sum \limits_{i = 1}^{m} \beta_{i} X_{i} + \mathop \sum \limits_{i = 1}^{m} \gamma_{i} Z_{i} + \mu } \right) = e^{{\beta_{0} }} \times e^{{\beta_{1} NATIONF}} \cdots \times e^{{\gamma_{1} GEN}} \cdots \times e^{\mu }$$
(2)

Research results and findings

Before conducting the Binary Logit regression, the explanatory variables needed to be tested for multicollinearity, considering that herders’ individual, household, and livestock production characteristics may have multicollinearity with the core independent variables. It is generally assumed that there is no significant multicollinearity between the explanatory variables at VIF ≤ 5. In this paper, the maximum value of the test result of multicollinearity of explanatory variables is 3.99 (VIF ≤ 5), the explanatory variables satisfy the principle of independence, and there is no severe cointegration problem.

After passing the multicollinearity test, to test the robustness of the model, control variables were introduced in the first step (model 1), nomadic sentiment and mutual aid sentiment were introduced in the second step (model 2), the proportion of animal husbandry income and balance of animal husbandry labour force was introduced in the third step (model 3), and all explanatory variables were introduced in the fourth step (model 4). The estimation results of the Binary Logit model are shown in Table 4. The regression results show that the change in the estimated coefficients of each explanatory variable is stable, and the significance level is improved after the core independent variables are introduced separately for estimation.The -2 Log-likelihood is often used as an index to compare the fit of different models, and the ‘-2 Log-likelihood’ of model 1, model 2, model 3 and model 4 changed from 392.815 to 322.300, showing a decreasing trend; Nagelkerke R2 is often used as a measure of the explanatory power of logistic regression models to the dependent variable. Nagelkerke R2 is often used as a measure of the explanatory ability of logistic regression, and ‘Nagelkerke R2’ changes from 0.407 to 0.556, showing an increasing trend and gradually approaching 1, which indicates that the model fit has been improved, and it has better model robustness.

Table 4 Estimation results of the Binary Logit model.
Full size table

Models 1–4 show that ‘nomadic sentiment’ and 'mutual aid sentiment,' 'the proportion of animal husbandry income’ and 'the proportion of animal husbandry labour force’ significantly and negatively affect herders’ willingness to transfer out of the pasture. The significance level of the control variables is the same after gradually introducing the core independent variables. The significance levels of the control variables are the same after gradually introducing the core independent variables. Based on the model’s results, the following is a detailed analysis.

National feelings and willingness to transfer out grassland

According to the estimation of model 2, the willingness of herdsmen to transfer out of the grassland decreases to 70% of its original level for every unit increase in nomadic sentiment. After introducing the social differentiation variable (model 4), herders’ willingness to transfer out of the grassland became 0.67 times the original, and the desire to move decreased slightly. Similarly, the influence of mutual aid sentiment on herders’ willingness to transfer out of grassland decreased from 0.80 times in model 2 to 0.78 times in model 4 after introducing the social differentiation variable, showing a slight decrease. It reflects the deep dependence of herders on their social status, cultural identity, and way of life. After adding the social differentiation variable, herders will consider the livelihood of their family members and the maintenance of their traditional pastoralist identity. Especially when the proportion of pastoral income and labour is high, transferring pastures may mean losing stable livelihood and social security support, so they are more inclined to stick to traditional pastoral production methods. Overall, the stronger the ethnic sentiment of herders, the lower the willingness to transfer pasture, and the more cautious they treat pasture transfer behaviour. That is, hypothesis 1 is verified.

The pastoral research found that since implementing the herders’ settlement housing project in 2009, under the vigorous promotion of governments at all levels, pastoralists have gradually moved to settlements and live-in houses of brick and wood or concrete structures. The traditional life and production of herders are affected to varying degrees. Some herders raise livestock in captivity and feed them by purchasing corn, forage or crop straw from agricultural areas, which increases costs. Still, the production is highly controllable, less affected by pasture drought and weather, and relatively less labourious and welcomed by herders, so the willingness of such herders to switch out of pasture is relatively high. However, some herders retain traditional nomadic production activities, with traditional felt houses or yurts replaced by shacked carts with better mobility. In summer and autumn, herders usually go into pasture to graze their livestock, return to the settlements in winter to feed their livestock with the stockpiled fodder, and then move to the meadow to graze their livestock when the lambs are born and grow strong enough in the spring. These herders’ nomadic sentiment and productive mutual aid sentiment are relatively high. The solid mutual aid sentiment increases the herders’ ability to adapt to resist abnormal climatic conditions and increases their confidence in livestock production. Thus, their willingness to turn out of pasture is relatively low.

Social differentiation and willingness to transfer out grassland

For every 1-unit increase in the proportion of animal husbandry income, the willingness to transfer out grassland is 0.011 times that of the original level (Model 3). It decreases to 0.010 times higher in Model 4 after including the nomadic and mutual aid sentiment variables. Household income is an essential factor constraining the scale of herders’ investment. As the proportion of herding income increases, herders’ investment in livestock production is willing to increase. Thus, the probability of switching out of pasture decreases significantly. It also confirms the research results of Zhao and Ba (2009) on investigating herders’ production and management behaviour in the Gannan pastoral area. For every 1-unit increase in the proportion of animal husbandry labour force, the willingness to switch out of grassland is 0.14 times that of the original level. It decreases to 0.11 times after including the ethnic sentiment variable. When the proportion of animal husbandry income and the proportion of the husbandry labour force are high, the herders’ economy and way of life are more homogeneous, so the adaptive ability to change in the external economic environment is weaker. After adding the variable of ethnic sentiment, the herders’ dependence on pasture is reflected in material, emotion, and spirit. This is because ethnic sentiment has a deep land complex, which makes it more difficult for herders to accept the cultural and emotional ‘separation’ brought about by the flow of pastureland22,23. That is, an increase in the proportion of animal husbandry income and the proportion of animal husbandry labour force significantly reduces the willingness to turn out of the pasture. Hypothesis 2 is tested.

Since the mid-1980s, when the "grass and livestock double contract" policy was implemented, collective pastures were divided and contracted to herd households, forming a family production and management pattern. However, as the population of herder families grows, the fragmentation brought by pasture division and subdivision increases the production cost of herders. It raises problems such as ecological degradation of pasture. The resource consumption and expenditure of herding families increase dramatically with the number of people, and against the background of the inability to enrich their sources of livelihood and the insufficient increase in the prices of livestock products, overloading overgrazing becomes the most helpless but effective direct choice for herders to increase their income, which naturally leads to pasture degradation. According to the estimation results of model 4, enhancing the added value of livestock production, reducing herders’ livelihood costs, enriching herders’ income structure, and promoting herders’ social differentiation (reducing the proportion of herding income/increasing the proportion of non-herding income) can, to a certain extent, increase herders’ willingness to switch out of the pasture. The effective allocation of pasture resources through pasture transfer realizes the scale operation of grassland animal husbandry and also helps to relieve the ecological pressure on the meadow.

Control variables and willingness to transfer out grassland

Individual characteristic: The estimation of models 1–4 showed that the effect of "being a local minority" and “education level” on the willingness to switch out of pasture was negative but did not pass the significance test. The reason for the non-significant ethnicity may be related to the distribution of the sample, as statistics show that 92% of the sample were local minority herders and only 8% were Han Chinese herders (see Table 1). The education level is not significant, indicating that there is not an inevitable trend of influence between herders’ willingness to switch out of pasture and education level, probably because herders’ love for grassland and grassland life remains the same due to the change in education level. The effect of “gender” and “age” on the willingness of pastoralists to switch out of pasture was positive and passed the significance test. Male pastoralists are, on average, 2.7 times more willing to turn out their pastures than females, which is related to the male society in pastoral areas. The pastoral research found that those who participated in the questionnaire responses were male members of the household. When asked to ask female family members, most requests were denied and indicated a lack of knowledge about household livestock production information. The average willingness of pastoralists aged [20,40] and [41,60] to switch out of pasture is 4.3 and 8.1 times higher than those aged 60 or older, respectively. Pastoral research found that the high willingness of pastoralists aged [20,40] to switch out is due to their preference for relatively easy “good jobs” in cities and their belief that the life of herding is hard and earning less. At the same time, a part of young herders believes that although herding is hard, if you know how to operate, expand sales channels, and enhance the added value of livestock products, operating livestock production still has great potential.

Family characteristics: The effects of "water supply status" and "electricity supply status" on herders’ willingness to switch out of pastures were positive, with the former having a non-significant effect and the latter having a significant impact. The main reason is that the area of above-ground runoff water in pasture areas has started to decrease. The underground runoff water is relatively stable, and herders are more inclined to exploit groundwater (drilling wells) to live and feed their livestock. At the same time, the cost required to use gasoline to generate electricity to drive the pumps is significantly higher than the cost of directly using electricity to drive the pumps. Thus, the effect of power supply conditions on the willingness of herders to switch out of pasture is more significant. The willingness of family members with the village and above public officials to transfer out of pasture is, on average, 4.9 times higher than that of herders in households without public officials. Herder families with public officials have better access to social information resources and relatively more channels and opportunities to change jobs and increase their income. Household health significantly and negatively influenced herders’ willingness to switch out of the pasture, with healthy households being, on average, 0.29 times more willing to switch out of pasture than sick households.

Characteristics of livestock production: “Livestock prospect” negatively influenced the willingness of herders to switch out of the pasture but did not pass the significance test. Theoretically, the better the herders’ future expectation of operating livestock, the lower the willingness to turn out pasture should be, but the regression results are not significant. The possible reason for this is that the research found that herders generally have reasonable expectations of livestock operations based on their expectations of the livestock industry or irrational behaviour based on national sentiment. “Livestock assets” and "number of livestock" significantly and negatively affected herders’ willingness to change out of pasture at 5% and 1% significant levels, respectively. The desire to switch out of pasture decreased to 0.66 times for each 1-unit increase in livestock assets, and the "number of livestock" decreased to 0.31 times for each 1-unit increase in livestock assets. For each unit increase in the "number of livestock," the willingness to switch out of pasture decreased to 0.31 times the original. When herders increase their investment in livestock production equipment, switching out of pasture means selling livestock production assets at a discount. Conversely, as capital investment increases, their willingness to switch out decreases. The increased number of livestock in herders’ households means that more pasture (forage) is needed to graze (raise) livestock. Thus, the willingness to turn out of pasture decreases significantly.

Further analysis

Endogeneity test

There may be an endogeneity problem between ethnic sentiment and pastoralists’ willingness to turn out of pasture, and this potential endogeneity may lead to biased estimates of regression results. On the one hand, there may be a reverse causality of pastoralists’ willingness to turn out of pasture on ethnic sentiment; on the other hand, ethnic sentiment and pastoralists’ willingness to turn out of pasture may be affected by unobserved omitted variables. Therefore, to alleviate the possible endogeneity problem of the model, based on the literature61,62, the instrumental variable method was adopted, and the 'average value of ethnic sentiment of other herding households of the same village (Gacha) with the same level of livestock assets’ was selected as the instrumental variable for the two-stage regression. This instrumental variable was chosen because the ethnic sentiment of herding households with the same level of livestock assets is convergent, and thus the ethnic sentiment of herding households is influenced by the ethnic sentiment of other herding households, which satisfies the correlation condition; the ethnic sentiment of other herding households in the same village (Gacha) does not have a direct effect on the willingness of herding households to turn out of the pasture, which satisfies the homogeneity condition.

The two-stage least squares method was applied to re-estimate the effect of ethnic sentiment on the willingness to turn out of pasture, and the results are shown in Table 5. The F-value of the first stage was 233.21, indicating that the weak instrumental variable test was passed. Column (1) of Table 5 shows that the coefficient of the instrumental variable is 0.334, which is statistically significant at the level of p = 0.01. It shows that there is a significant correlation between 'the average value of ethnic sentiment of other herding households in the same village (Gacha) with the same level of livestock assets’ and ethnic sentiment, which verifies the validity of the instrumental variable; Column (2) shows that after alleviating the endogeneity problem, the negative impact of ethnic sentiment on herders’ willingness to switch to new farms is still significant, and therefore the previous analyses are credible.

Table 5 Regression results of ethnic sentiment on pastoralists’ willingness to turn out of pasture based on instrumental variables approach.
Full size table

Robustness check

In this paper, we choose to test the robustness of the benchmark regression results by replacing the regression model and changing the sample size, respectively. Table 6 reports the robustness test results of the benchmark regression. Firstly, replace the econometric model. Considering that different model selections may impact the analysis results, both the Binary Probit model and the Binary Logit model can effectively address binary choice problems. The difference between them lies in that the former follows a normal distribution. In contrast, the latter follows a logistic distribution, leading to slight differences in the coefficients estimated by the two models. However, this does not affect the research conclusions. Therefore, following the approach of Liu et al.63, the Binary Probit model is selected to test the robustness of the regression results obtained from the Binary Logit model. Secondly, alter the sample size. Referring to the research of Brodeur et al.64, we randomly select 80% of the samples to conduct regression analysis again, to exclude the interference of P-hacking behavior on the benchmark regression. Based on the robustness test results, it is found that the regression results are basically consistent with the previously mentioned regression results. The significance and signs of the explanatory variables and the explained variable have not changed significantly, indicating that the benchmark regression results exhibit good robustness.

Table 6 The robustness test results of benchmark regression.
Full size table

Conclusions and suggestions

Conclusions

  • (1) Nomadic and mutual aid sentiments significantly and negatively influence herders’ willingness to switch out of the pasture. The stronger the herders’ ethnic sentiment, the lower their willingness to transfer out of pastureland, and the more unfavorable it is to the development of the pasture transfer market, revealing the centrality of cultural and socio-emotional factors in herders’ decision-making. Traditionally, pastoralists’ pasture transfer problems are analyzed mainly from the economic perspective, focusing on improving land use efficiency and economic returns through market-oriented reforms4. However, the introduction of ethnic sentiment has expanded the sociological dimension of the issue, suggesting that pastoralists are not only economically rational individuals but are also affected by ethnic identity, cultural inheritance, and social connectivity. This interdisciplinary perspective encourages scholars to pay more attention to the complexity of the 'culture-society-economy’ interweaving, further promoting the direction of multidisciplinary research. As early as 1972, Smith and Martin reported that Arizona ranchers, because of their love of the land and rural values, were reluctant to turn out pastures even when market prices far exceeded livestock values. In contrast, ranchers with lower ethnic sentiments focused on transferring pastureland, a phenomenon remarkably similar to that on rangelands in the western United States. Ranchers transfer pastures to ‘amenity’ ranchers65, who are more concerned with the socio-ecological, cultural, and landscape values of the pasture66, and ranchers maintain the ecological landscape of their pastures and can earn a corresponding income after engaging in pasture transfer behaviour67,68. With the continuous economic and social development of pastoral areas and the implementation of national ecological policies, herders are gradually accepting modern animal husbandry techniques and business methods. Led by the government and large local households, the inhibiting effect of ethnic sentiment on herders’ participation in pastureland transfer has been decreasing, and different degrees of differentiation have occurred between herding areas and farming areas.

  • (2) The proportion of animal husbandry income and the proportion of animal husbandry labour force significantly and negatively affect pastoralists’ willingness to switch out of the pasture. It means that the lower the proportion of animal husbandry income and the proportion of animal husbandry labour force, the higher the willingness of pastoralists to transfer out of the pasture and the more favourable to the development of the pasture flow market. Pastoralists’ income sources and labour resource dependence directly affect their attitudes and behaviour towards pasture flow. From an economic point of view, the more homogeneous the herders’ income structure, the more important the role of pasture as a means of production, and the more they tend to prioritize the long-term impact of pasture flow on their livelihoods. Especially in traditional pastoral areas where herders rely on many labourers, pasture turnover may directly lead to changes in the household labour structure and even make family members face the pressure of employment transition and social role change. Nori52 shows that pastoralist systems worldwide are undergoing rapid change, with the low population density of pastoral resources gradually being valued by external investors. Since the division of pastureland into households, pastoralists have been renting livestock and pasture to each other to form a flow, and social relations between pastoralists have been monetised as they become market actors31. The structure of the household economy is influenced by the distribution of household labour and the share of household income, and pastoralists support the household economy through different degrees of differentiation. Social differentiation in pastoral areas has become an intrinsic driving force for pastoralists’ pasture flow and livelihood strategy choices.

  • (3) The heterogeneous characteristics of herders’ household endowment resources are the basis for enriching the mode of livestock production and operation or business structure in pastoral areas and one of the driving forces for shifting herders’ livelihood strategies. Furuseth and Pierce69 suggest that if one wants to influence the form of land transfers, it is necessary to understand the leading decision makers and their relationships. In particular, differences based on herders’ individual, household, and livestock production characteristics affect herders’ pasture transfer behaviour and significantly affect livestock production decisions, industrial transfer, and other behaviours.

Suggestions

Transformation of thinking and innovation of grassland livestock management is the initiative to implement the rural revitalization strategy in pastoral areas in the new period. Xinjiang and Inner Mongolia have achieved initial results in their attempts to carry out cooperative share and community-led pasture co-management through pasture transfer. Effectively matching pasture resources is a meaningful way to promote large-scale grassland animal husbandry operations. Based on the above findings, several recommendations are made:

  • (1)Taking advantage of the situation and giving full play to the positive role of pastoralist ethnic sentiment in implementing the strategy for revitalizing pastoralist villages. In pastoral areas, the negative impact of ethnic sentiment on willingness to transfer challenges the traditional animal husbandry model. The traditional 'pasture transfer = economic development’ formula is no longer valid. This theoretical challenge requires policymakers to fully consider local cultural and historical backgrounds when promoting pasture transfer and to understand and respect herders’ ethnic sentiments as a prerequisite for promoting pasture transfer and modernization in pastoral areas. Therefore, the government should explore the mode of collective pasture transfer and ensure that herders can maintain their ethnic sentiment and cultural identity while participating in the modernization process and enjoying the economic benefits through the method of 'common ownership and common management.' In addition, flexible pasture transfer models should be developed according to herders’ different needs and emotions. For example, a 'contracting + cooperation’ approach should be implemented, allowing land to be transferred to enterprises or cooperatives while ensuring that herders can still participate in pasture management and operation. It will help to effectively connect the production and investment needs between herders and enterprises, enrich the livestock business model, and promote the long-term stable development of the livestock industry.

  • (2) Enhance herders’ employability, enrich income channels, and stimulate the endogenous power of herders’ livelihood development through social differentiation. Subject to the constraints of their endowments, herders have different income levels and labour dependence, which constrain them in formulating behavioural strategies for turning out pastureland. It is recommended that differentiated pasture transfer policies be designed; for herders with a higher proportion of animal husbandry income and a greater dependence on labour, a transition period should be introduced, or short-term economic support should be provided to help herders gradually adapt to the transition. For herders with diversified sources of income and lesser dependence on labour, participation in the transfer of pasture should be encouraged, with more flexible and market-oriented conditions for the transfer of pasture being provided. On the one hand, herders’ vocational skills and income diversification capacity should be improved through enhanced training in livestock production skills and professional and technical training, bridging the workforce quality needs for developing high-quality agro-industries and optimizing the structure of the family labour force. On the other hand, it has assisted herding households with little grassland and low incomes in improving their family incomes by transferring grassland, thereby promoting the sustainable development of grassland resources in pastoral areas.

Methods

Data source: The data we used came from related studies conducted by the group in October 2018 and August 2022 when they traveled to pastoral flag counties in Inner Mongolia and Xinjiang. To ensure the validity and accuracy of the survey, a pre-survey was conducted before the formal study. The questionnaire survey method was used, and the questionnaire content that consumers did not understand was answered accordingly during the interview process to ensure that the questionnaire was credible reliable, and in line with the real situation of the respondents. After the pre-survey, we generally reflected on the research process, some of the concepts were difficult to understand the number of questions was too large to correct the problem, and some of the questions were not close to the core content of the study.

Data inclusion: The herders in the study area are mostly ethnic minorities, and most of them are middle-aged and old. To ensure the smooth progress of the study, Mongolian and Kazakh university students were recruited and trained on the questionnaire in the school before the study, and “face-to-face” interviews were conducted, in which the members of the research team asked questions in a spoken form, and Kazakh students translated. The members of the research team recorded the questions.

A statement: It is confirmed that all investigations were performed in accordance with relevant guidelines and regulations. This investigation was supported by the National Social Science Foundation of China, the Scientific Research Project for Universities of Xinjiang Education Department, and the Postgraduate Innovation Programme for the Autonomous Region 2024. It was endorsed by and licensed by the project. The data of this study are authentic and reliable, and it is confirmed that informed consent was obtained from all legal guardians and all the participants.

It is confirmed that all methods were carried out in accordance with relevant guidelines and regulations.

It is confirmed that all investigation protocols were approved by a named institutional and licensing committee (College of Economics and Management, Xinjiang Agricultural University).