Can Social Capital Help the Youngers Rent Land?—A Case of Pastoral Areas in Inner Mongolia, P.R. China

Abstract

Ensuring the youngers obtain access to land resources is important to help them earn a living and sustain agriculture. By applying a probit model to a comprehensive dataset with 422 sampled households from Inner Mongolia, the results of this study showed that social capital significantly improved the possibility of herders renting land. However, compared with the middle-aged group, the youngers’ social capital is significantly weaker in helping them rent land. The findings and policy implications proposed by this study are expected to narrow the intergenerational difference in social capital and promote the transfer of land between generations, which will help the youngers obtain access to land resources, enhance their sustainable livelihood, and facilitate the sustainability of agriculture in the future.

1. Introduction

Youth unemployment and migration are growing challenges that need more political attention in many countries [1]. This is particularly true for youths in agricultural and pastoral areas where land resources are the basis of production. Ready access to land is foundational for improving farmers’ and herders’ living standards [2], especially for the youngers under 35 years old in Inner Mongolia, P.R. China, who were born too late to take advantage of land contracts from villages during land distribution. Due to a lack of nongrazing skills, they cannot seek off-farm jobs competitively with their counterparts from other areas. Thus, the youngers are highly vulnerable economically. Engaging in animal husbandry in their hometown becomes the primary occupational option for some youngers. In addition, the youngers’ engagement is also needed to sustain agriculture in the future, similar to conditions in Ethiopia [3]. Land access is a barrier for youth entry into agriculture, as observed in Africa [4]. The youngers were not allocated village land in the 1980s and 1990s. They can obtain land resources in only two ways: inheriting land from parents or renting from the land market. Since inherited land is often too small (usually less than 20 ha each plot) and fragmented (more than three scattered plots) to sustain a family [5], participating in the land rental market is thus becoming the most important means for them to obtain land [6]. Although rental markets may cause land degradation [7,8], they have been found to increase the technical efficiency of herders involved in the markets [9].

Despite their need, it is difficult for the youngers to rent land. One explanation for this is that Chinese society depends greatly on “guanxi” (relationships) [10], both in agricultural and pastoral areas [11]. Social capital based on kinship and geopolitical location has a significant impact on rural land transfer [12]. With these considerations in mind, this study addressed two issues related to the youngers and the land rental market: (1) What was the degree of the younger participation in the land rental market? and (2) Did the effect of the youngers’ social capital on their land rental decisions differ from that of the older generations?

There are several studies on land rental markets from the standpoint of social capital both globally and in China. For example, Robison et al. [13] used an economic model to evaluate land transactions with strangers and acquaintances and discovered that social capital strongly affected land transaction pricing. Holden and Ghebru [12] used data from northern Ethiopia to investigate the influence of kinship on land rental market participation in terms of transaction costs. Tenant access was found to be less regulated in areas with a high proportion of kinship contracts. According to Niu [14], kinship impacts land transfer decisions made by farmers in rural China. Chen and Wang [15] discovered that social capital is vital in encouraging land transfer in rural China because high transaction costs induced by information asymmetry prevent farmers from entering the market [16,17]. According to Chen et al. [18], social capital lowers transaction costs and assists participants in renting land resources [19,20].

Studies examining the impact of social capital on the youngers’ participation in land rental markets are rare. The youngers have limited potential to obtain land [3], and social capital can greatly affect an individual’s land rental decision [15]. A good understanding of how social capital influences the youngers’ participation in land rental markets is needed to help them obtain land resources, reduce their economic vulnerability, and sustain agricultural development.

This study analyzed data gathered from 422 households in 2011–2012 from Inner Mongolia, P.R. China, where the extent of pastoral lands is the largest in the country, accounting for approximately one-quarter of China’s total rangeland. Most households received livestock in 1983–1984, and grasslands were distributed to individuals in the mid-1990s [9]. The interaction between age and social capital was investigated in terms of the youngers’ engagement in the land rental market. To the best of our knowledge, this is the first study based on such a large survey that looks at the impact of social capital on the youngers’ participation in the land leasing market in China’s pastoral areas. This is most likely due to the difficulty of gathering data from large, sparsely inhabited grassland areas. The main purpose of this study is to test the effect of social capital on herders’ involvement in the grassland rental market and to explore whether this effect has intergeneration heterogeneity. The findings of this paper are meant to provide policy suggestions to help the youngers obtain access to land resources to sustain an agricultural lifestyle in the future. We believe that the implications of this study can be applied to similar regions and countries to promote land policy interventions that ensure the livelihoods of the youngers and sustain agriculture. The remaining part of the paper is organized as follows: Section 2 introduces the changes in the youngers’ access to land. Section 3 presents the theory and hypotheses. Section 4 elaborates the materials and methods. Section 5 and Section 6 present the results and discussion, respectively. Section 7 concludes the study.

2. Changes in the Youngers’ Access to Land at the Study Site

2.1. Traditional Methods of Land Acquisition

Traditionally, lands in Inner Mongolia’s pastoral regions have been handled communally [21]. Herders relocated to areas with abundant feed and water. The extreme disparities in grazing patterns and the establishment of regulations for unsettled land were seasonally influenced by meteorological circumstances. Low rainfall and restricted land production required herders to adopt transhumance to decrease grazing intensity on stressed areas and provide pastures with enough time to regenerate. Herders would co-manage the land and help one another amid frequently recurring disasters. The youngers often acquired land access by co-managing lands with the older generations, gaining grazing expertise in the process.

The nomadic system was created to ensure that all individuals had access to community land resources. The traditional nomadic patterns have been criticized as inefficient and unclear borders resulted in overgrazing and grassland damage [22,23]. Clarifying property rights and borders was suggested as a better strategy to manage grasslands.

2.2. Reform of the Land Tenure System in Pastoral Areas

Because the Household Responsibility System (HRS) was regarded as effective in agricultural regions [24], the HRS grassland tenure reform was introduced in pastoral areas some years later. The grassland contracting system (GCS) was implemented in China’s pastoral districts in the 1980s [25] to increase animal husbandry efficiency and avoid “the tragedy of the commons” [23]. This approach was put in place in two stages: livestock distribution and land distribution. The latter was separated into two phases: the first began in the 1980s, and the second in the mid-1990s, when lands were allocated to individual households in 30-year contracts [26].

Under the GCS, land cannot be redistributed administratively. Those who did not receive land during GCS do not have the chance to contract land directly from local villages. As a result, many herders experience feed shortages resulting from highly restricted land holdings. As a result of this requirement, a land rental market has emerged to assist herders in gaining access to land and reallocating land resources in pastoral regions. Although some researchers have argued that the land rental market in pastoral areas may result in severe land degradation [7], it is generally considered an efficient way to reallocate land resources and increase production efficiency [9].

Land rental market regulations in Inner Mongolia were described in detail by Tan et al. [2]. Notably, policies designed to help the youngers obtain access to land through the rental market are absent from these policies.

2.3. Comparison of Access Routes to Land Rentals for the Youngers

Because land trafficking is forbidden in China, herders may access land in pastoral regions through one of only three channels. The first is government land allotment or administrative redistribution. Administrative land redistribution is a partial reallocation of land resources based on the land contracting system, with the goal of guaranteeing equity. People under the age of 35, however, missed the first land distribution in the 1980s, and the government does not have plans to transfer land to them in the near future. The second channel is through family inheritance (parents or grandparents). Custom dictates that land inheritance must be evenly distributed among sons (and daughters in some areas, e.g., the Tibetan region). This makes land holdings more fragmented, and the land obtained from an inheritance is often too small to sustain a family. The third channel is via the land rental market. Compared to administrative redistribution and inheritance, the rental market is an efficient and timely way for the youngers to obtain land [27]. For these reasons, the land rental market has grown more significant in reallocating land resources under the existing tenure structure, particularly for the younger generation.

3. Theory and Hypotheses

3.1. Who Are the Youngers?

There is no universally agreed-upon definition of the youngers. According to current criteria, the International Labor Organization considers everyone above the age of 45 to be elderly. In the United States, a young farmer is defined as one who is under the age of 35 by the Farm Bureau and Farm Credit organizations. The EU’s Common Agricultural Policy (CAP) also regards farmers under 35 as young farmers. People often prefer to refer to the next generation of farmers as “young farmers.” Although there is no consistent definition of “young herders” or “young farmers” in China, the common definition of “new generation farmers” is those born after 1980 [28]. New generation farmers inherit land from their parents and/or rent land. Some scholars define “new generation farmers” as agricultural laborers with strong physical capability who are between 16 and 31 years of age [29]. For this study, according to the mainstream definition, we defined the youngers as individuals who are under 35 years old, and the majority of this age cohort missed land contracts from villages.

3.2. Why the Youngers?

Will the current youngers be the major force sustaining agriculture in the future? Goldman et al. [30] interviewed herders in Mongolia and asked them, “Do you want your children to grow up to be herders?” All replied, “No! I do not want my children to be herders; it is too difficult, and the future is uncertain.” A similar situation exists in the pastoral areas of P. R. China, including Inner Mongolia. When we asked the herders in this area, “If all the youngers do not come back (after graduation or from off-farm activities) to pastoral areas, who will be responsible for grazing in the future? And what about the future of the pastoral area?” Most of the herders had no answer. We also asked, “How many young herders come back to the pastoral areas every year?” A village head replied, “Approximately 4–5 young herders come back to the village each year, and if they had a way to obtain more grassland, I believe more young herders would come back to develop animal husbandry in the pastoral areas.” In fact, we found that some youths wanted to engage in animal husbandry, despite the difficulty of herding: “It is difficult for me to find a job in the town and the salary is quite low. I have inherited some grasslands from my father in the pastoral area and eventually I will go back to graze livestock,” remarked a 28-year-old herder in the Sonid Right Banner in Inner Mongolia.

The youngers at the study site face difficulties when seeking jobs in urban areas compared to others with nongrazing skills. Additionally, they lack land resources compared to the elder generations. Thus, they are now a vulnerable group, and if more youths give up farming, it will be difficult to predict what will happen in the future [30]. Ensuring the youngers are able to rent land can help them reduce their economic vulnerability and promote the sustainable development of agriculture.

3.3. Social Capital, Transaction Cost, and Land Rentals

Social capital is a positive outcome of human interaction that can be tangible or intangible and can include useful information, novel ideas, and future opportunities [31]. We define social capital in this research as trust, reciprocity norms, and networks among herders that could be used for individual or communal profit [32,33]. Herder social capital has been changed by the transitions in the grassland tenure system. Before the GCS, pastures were managed by communes, and herders in the same communes shared livelihood assets. After the GCS, these assets were allocated to individual households in terms of family size and flock size. Such privatization discouraged cooperation and weakened the herders’ social capital [26].

With an imperfect grassland rental market, transaction costs are significant due to information asymmetry. This raises the land rental market threshold [34]. Social capital, which acts as a bridge for information and a foundation of trust, plays an important role in lowering transaction costs in two ways. To start, it promotes the exchange of information. Reduced information asymmetry can aid in the reduction of transaction costs. Interpersonal connections are the manifestation of social capital, and they can be classified as strong or weak [35]. Strong relationships aid in the acquisition of more difficult-to-obtain information or resources, whereas weak relationships aid in the acquisition of fresh, nonrepetitive information or resources. With increased social capital, the breadth of a herder’s intrapersonal world expands and transaction costs decrease, potentially decreasing the entry hurdle into the grassland leasing market.

Second, social capital lowers supervision costs because reciprocity rules minimize moral hazard and opportunism [15]. The process of formalizing the land transfer system is still in its early stages, with many land transfer transactions relying solely on oral contracts [36]. If there are breaches in oral contracts, a formal system for grievances will result in greater expenses. Because people who violate contracts face punishment from within the social network, establishing reciprocity standards is more successful in avoiding contract breaches. This reduces moral hazard and the likelihood of opportunism, lowering monitoring expenses.

We suggest the first hypothesis based on the analysis:

Hypothesis 1 (H1).

Social capital increases herders’ participation in the land rental market.

3.4. Differences in Social Capital and Land Rental Decisions between Generations

Many studies have found that social capital lowers transaction costs [18] and makes land transfer easier [15]. Fewer studies, however, have found that interpersonal networks exhibit age variations and clustering aspects [37]. Humans build social networks based on their interests and activities [38], which subsequently serve as the foundation for social capital. This is a more typical phenomenon in the information technology era [39]. These activities, as well as the ensuing social environment, vary among generations [40]. Social contacts for the youngers are mainly individuals who have the same or comparable worldview and beliefs.

The senior generations and the youngers built their social networks with their peers, resulting in a “strong relationship” among peers but a “poor relationship” between the two generations (Figure 1). It was difficult for the youngers to rent land from older people since the older generations had more land and were more inclined to lend grassland to other members of their same age cohort to minimize the lessors’ risk [41]. However, because the younger had a low supply of land resources, it was impossible for one young herder to rent land from another based solely on their social network.

We suggest the second hypothesis based on this analysis:

Hypothesis 2 (H2).

The youngers’ social capital does not affect their rental decisions as much as elder generations’ social capital.

4. Materials and Methods

4.1. Sampling and Data Collection

The data used were gathered from Hulun Buir and Xilin Gol—two grassland-dominant leagues of Inner Mongolia (Figure 2). With one-third of the total grassland area in Inner Mongolia, these two leagues account for one-fourth of total livestock in the region. The two leagues are also representative in view of their diversified grasslands, including meadow steppe, steppe, desert steppe, and desert vegetation.

In 2011 and 2012, our team conducted face-to-face interviews using a standardized household questionnaire that included questions on herder characteristics, grassland renting decisions, credit status, risk preference, and social capital. Four banners/cities in Hulun Buir were examined, as were eight banners/cities in Xilin Gol. From each banner, two to seven townships were chosen, three to four villages from each township were visited, and four to eight herders from each village were picked at random to be interviewed. After training the interviewers and translators and conducting some exploratory surveys around Hailar and Xilin Hot—the capital cities of these two leagues—the team was divided into three groups with five to seven members in each group. Led by experienced staff and a local guide/translator, each group was responsible for surveying herders in several banners. It took two to three hours to interview each head of household. In total, 430 households were interviewed, and 422 samples were used after data cleaning (

Table 1. Survey sample groups based on the number of livestock normalized by SSU *.

Grouping Criteria (SSU)	No. of Herders	Ratio	Mean (SSU)
<200	144	34%	109
200–500	148	35%	325
>500	130	31%	955

Note: Data source: from authors’ field survey. * SSU means standardized sheep units, 1 sheep = 1 SSU; 1 goat = 0.9 SSU; 1 cattle = 5 SSU; 1 horse = 6 SSU; 1 camel = 7 SSU.

).

4.2. Model Specification

This question was investigated using the probit model: “Can social capital influence herders’ participation in the grassland rental market?” The model is a statistical regression with a normal distribution. The dependent variable is a binary variable having a value of either 0 or 1. Equation (1) is used to test Hypothesis 1, namely, the effects of social capital on herders’ renting decisions. “φ(∙)” is a cumulative distribution function. “y” represents whether herders participated in the rental market: if herders rented in lands,” y = 1”; otherwise,” y = 0”. “${\mathrm{x}}_{\mathrm{i}}$” is a vector of factors that may influence herders’ grassland rental decisions. We generated an age dummy by dividing the herders into 3 groups, namely, young herders ($\le$35), middle-aged herders (36–59), and older herders ($\ge$60).

$$\begin{gathered} \mathrm{p}\left(\mathrm{y}=1|\mathrm{x}\right)={{\displaystyle \int }}_{-\infty }^{\mathrm{x}\mathsf{\alpha }}\mathsf{\phi }\left(\mathrm{t}\right)\mathrm{dt} \\ =\varnothing \left({\mathsf{\alpha }}_0{+\mathsf{\alpha }}_1·dum\_age(age \le { 35)+\mathsf{\alpha }}_2·dum\_age(age \ge { 60)+\mathsf{\alpha }}_3·social\_capital+{{\displaystyle \sum }}_3^{\mathrm{n}}{\mathsf{\alpha }}_{\mathrm{i}}{\mathrm{x}}_{\mathrm{i}}{+\mathsf{\epsilon }}_{\mathrm{i}}\right) \end{gathered}$$

(1)

To further test H2, an interaction term of age dummy and social capital was generated, namely, $social\_capital·dum\_age\left(·\right)$. It is an index function, e.g., when age is less than 35 years, $dum\_age\left(age\le 35\right)=1$; otherwise, $dum\_age\left(age\le 35\right)=0$. Similarly, if age is older than 60 years, $dum\_age\left(age\ge 60\right)=1$; otherwise, $dum\_age\left(age\ge 60\right)=0$. To avoid “dummy variable traps”, herders aged between 36 and 59 years old were selected as the benchmark group. This is depicted in Equation (2).

$$\begin{gathered} P(y=1|x)={{\displaystyle \int }}_{-}^{x\alpha }\phi \left(t\right)dt=\varphi ({\beta }_0+{\beta }_1·dum\_age(age\le 35) \\ +{\beta }_2·du{m}_{age\left(age\ge 60\right)}+{\beta }_3·socia{l}_{capital+{{\displaystyle \sum }}_3^{\mathrm{n}}{\mathsf{\alpha }}_{\mathrm{i}}{\mathrm{x}}_{\mathrm{i}}{+\mathsf{\epsilon }}_{\mathrm{i}}} \\ +{\beta }_4·social\_capital·dum\_age\left(age\le 35\right)+{\beta }_5·social\_capital·dum\_age(age\ge 60) \end{gathered}$$

(2)

4.3. Measurement of Social Capital

We distinguished social capital into three dimensions (

Table 2. Measurement of herders’ social capital.

Dimension	Question/Statement	Mean	Std. Err	Note
Social Network	Do you belong to any political parties, unions, clubs, or other organizations? (S1)	0.39	0.49	1 = yes 0 = no
	It is simple to obtain animal husbandry knowledge and technology (S2)	3.1	1.17	Likert-like scale
Trust	I have a large number of good pals (S3)	2.1	0.7	Likert-like scale
	I think everyone is trustworthy (S4)	6.69	2.37	0–10 score
Reciprocity norms	In recent years, average relationship expenditure including weddings and burials (S5)	2.86	1.13	1000 yuan
	Would you be willing to contribute to a project that might benefit your community? (S6)	1.48	0.87	Likert-like scale

Data source: Inner Mongolia field survey.

), i.e., social networks, trust, and reciprocity norms. Each dimension consisted of several questions. Questions S2, S3, and S6 were answered on a 5-point Likert-like scale, with 1 = fully agree, 2 = agree, 3 = not sure, 4 = disagree, and 5 = entirely disagree. The answer to S4 was scored between 0 and 10, where 0 = totally disagree, and 10 = fully agree.

The social network dimension was measured by questions S1 and S2. S1 reflected herders’ participation in other parties, unions, clubs, and groups, among which the role of blood and kinship is also reflected. The blood- and kinship-based group members are usually family members and/or close relatives. Normally, they frequently communicate through WeChat, phone communications, and family celebrations (e.g., wedding parties, birthday parties, and funeral receptions). These frequent communications among blood or kinship groups can reinforce their relationship and increase social capital. S2 was about the ability of herders to obtain information.

The trust dimension was represented by S3 and S4. These two questions measured the herders’ trust in others. The reciprocity norms were measured by S5 and S6. S5 measured relationship spending. In China, when a family holds a wedding or funeral reception, cash gifts from others are used to express congratulations to the bride and groom or express appreciation and gratitude for the deceased.

According to the results depicted in

Table 3. Variable explanations and descriptive statistics.

Variable		Unit	Mean	Std. Dev.	Note
Dependent variable	Rent in land	-	0.32	0.47	1 = yes, 0 = no
Social capital	Social capital index	-	9.43	2.07	Calculated based on Table 2
Herder’s characteristics	dum_age	-	1.87	0.56	Age dummy of household head 1 = younger than 35, 2 = 35~60, 3 = older than 60
	Education	-	2.91	1.04	1 = illiteracy, 2 = primary school, 3 = junior school, 4 = high school and above
	Labor	number	1.96	1.03	Available labor in household
Financial capital	Non livestock income ratio	%	0.26	0.29	Non-livestock income/ total income
	Investment	CNY 1000	18.35	38.34	Value of lawn-mower, fence, sheds, wells, veterinary services
	Livestock	SSU	311.19	332.8	Stocked animals in SSU
Natural capital	Hayfield	ha	103.55	147.96	Area of contracted (inherited) hayfield
	Pasture	ha	211.18	294.16	Area of contracted (inherited) pastureland
Policy variables and Regional characteristics	Forbidden *	-	0.18	0.39	1 = yes, 0 = no
	Balance *	-	0.51	0.5	1 = yes, 0 = no
	Location	-	0.52	0.5	1 = Xilin Gol, 0 = Hulun Buir

* Herders were not permitted to graze animals on “grazing-forbidden” plots under previous legislation. The “grassland-livestock balance” strategy required herders to graze without exceeding the carrying capacity of grassland. Data source: field survey in Inner Mongolia.

, we used the factor analysis method to calculate an index of social capital. In Stata 12.0, Kaiser–Meyer–Olkin tests (KMO) were employed to investigate the fitting effect. The KMO was 0.85, which is greater than the crucial value of 0.6, indicating that the factor analysis approach could be applied for the specified variables and that the fitting effect was good. We found three primary common variables based on the results of KMO and rotation, namely, F1, F2, and F3. The overall variance contribution was 90.9%, with F1, including S1 and S2, contributing 45.5% to the variance. People who joined political parties, unions, clubs, or other organizations (see S1) frequently had a larger social network and thus were more likely to obtain information and technology, as social networks are the main way for herders to gain these advantages. F2 consisted of S3 and S4, which together contributed 35.0% to the total variance. A higher trust level meant that the herder had more friends (see S3). Similarly, a higher trust level meant that herders thought everyone was more trustworthy (see S4). F3 consisted of S5 and S6, with a contribution of 10.4% to the variance. A higher degree of reciprocity value norms means more money spent on maintaining the relationship and more willingness to make contributions to the village (see S5 and S6). According to the variance contribution obtained, we used a weighted combination method and calculated the variable $social\_capital as (F1*45.5+F2*35.0+F3*10.4)/90.9$.

5. Results

5.1. Descriptive Statistical Results

Table 3 and

Table 4. Descriptive statistics of variables.

Variables	Option	Percentage (%)	Variables	Option	Percentage (%)
Education	1 = illiteracy	9.5	S2 in social network	1 = completely agree	7.7
	2 = primary school	30.3		2 = agree	30.8
	3 = junior school	19.7		3 = not sure	10.3
	4 = high school and above	40.5		4 = disagree	36.8
Chinese level	1 = cannot speak Chinese	11.4		5 = completely disagree	14.5
	2 = can speak a little Chinese	37.2	S3 in trust	1 = completely agree	16.5
	3 = speak Chinese smoothly	51.4		2 = agree	57.4
dum_age	age $\le$ 35	23.2		3 = not sure	25.2
	35 < age < 60	66.8		4 = disagree	0.7
	age $\ge$ 60	10.0		5 = completely disagree	0.2

show the descriptive statistical results of “y” and “${\mathrm{x}}_{\mathrm{i}}$” in Equations (1) and (2). The mean value of “y” represents that 32% of the sample rents land. “${\mathrm{x}}_{\mathrm{i}}$” includes five categories: herder characteristics, natural capital, financial capital, social capital, and policy variables and regional characteristics. Specifically, sample herders younger than 35 years old, between 36 and 59, and above 60 accounted for 23.2%, 66.8%, and 10.0%, respectively. Herders’ education level was not high, with approximately 10% illiterate, 30% with just primary school, approximately 20% with just junior school, and the remaining 40% at high school or above. We found that 11.8% of herders could not speak Chinese, 38.6% could speak a little, and 49.6% of herders could speak fluent Chinese. Social capital ranged from 1 to 18, with a mean value of 9.43 and a standard deviation of 2.07. Table 4 shows the statistics of S2 in social networks and S3 in trust in a more detailed way.

Table 5. Herders’ grassland rental market participation related information.

	Age	$\le$35	36–59	$\ge$60
Variable
Numbers of herders		98	282	42
Percentage of total (%)		23.22	66.82	9.96
Numbers of rented in herders		36	95	6
Percentage of renters (%)		36.73	33.69	14.29
Rental price (CNY/ha.month)		7.88	7.08	6.25
Social capital index		8.95	9.55 (0.016 **)	9.77 (0.047 **)

Data source: field survey in Inner Mongolia. Note: The results of t tests in parentheses show that both differences in the social capital index between younger herders and the middle-aged group/elder group are significant at the 5% level. ** p < 0.05.

compares the information on the participation of young herders in the grassland rental market with that of their elder counterparts. Among the total herders who participated in grassland rental markets, herders younger than 35 years old accounted for less than one-fourth. Moreover, these young herders faced significantly higher rental prices (7.88 CNY/ha.month) than their elder counterparts (7.08 CNY/ha.month) (Table 5). At the same time, the youngers have significantly lower social capital than the elder herders (Table 5). According to our field surveys, on average, the costs of raising a family with three people meant a minimum of 200 SSU in total wealth, and an SSU needed approximately 1.0 hectares in the study areas. This means that, to make a living, young herders needed to rent approximately 205 ha of grassland. They had to pay more than 164 CNY/month, or nearly 2000 CNY/year more than their middle-aged counterparts. This is a significant amount for the young herders, as the annual per capita income in Xilin Gol in 2011 was approximately CNY 6800. Although some young herders complained about higher rental prices, participating in the rental market was a major means of access to grassland for them.

5.2. Model Results

Two probit models were estimated using Stata 12.0. Model 1 is for H1 with the age of the head of household as a dummy variable. Model 2 is for H2 with an interaction term of age dummy and social capital.

Table 6. Results of the Probit model evaluation with robust and marginal effects.

Variables	Model 1		Model 2
	Probit	Marginal Effect	Probit	Marginal Effect
dum_age ($\le$35)	−0.662 **	−0.178 ***	−1.148 *	−0.25 ***
	(−2.41)	(−3.13)	(−1.65)	(−2.96)
dum_age (age $\ge$ 60)	0.050	0.016	2.607 **	0.807 ***
	(0.31)	(0.31)	(1.81)	(3.38)
Social capital	0.085 *	0.028 *	0.245 *	0.079 **
	(1.89)	(1.91)	(1.88)	(1.97)
Social capital * dum_age ($\le$35)	-	-	−0.281 *	−0.091 **
	-	-	(−1.94)	(−2.03)
Social capital * dum_age (age $\ge$ 60)	-	-	0.049	0.016
	-	-	(0.74)	(0.74)
Chinese level	0.326 ***	0.107 ***	0.310 ***	0.100 ***
	(2.75)	(2.78)	(2.60)	(2.58)
Ratio of non-livestock income	−1.123 ***	−0.371 ***	−1.192 ***	−0.385 ***
	(−3.48)	(−3.53)	(−3.52)	(−3.64)
Education	0.038	0.012	0.053	0.017
	(0.53)	(0.54)	(0.76)	(0.76)
Labor	0.023	0.008	−0.004	−0.001
	(0.30)	(0.30)	(−0.05)	(0.76)
Investment	0.027	0.009	0.025	0.008
	(0.89)	(0.89)	(0.79)	(0.79)
Livestock	0.146 **	0.048 **	0.160 **	0.052 **
	(2.33)	(2.39)	(2.50)	(2.55)
Hayfield	−0.158 ***	−0.052 ***	−0.157 ***	−0.051 ***
	(−3.45)	(3.51)	(−3.36)	(−3.36)
Pasture	0.008	0.003	−0.003	−0.001
	(0.21)	(0.21)	(−0.07)	(−0.07)
Forbidden	0.28	0.097	0.289	0.099
	(1.43)	(1.37)	(1.48)	(1.42)
Balance	−0.18	−0.059	−0.131	−0.042
	(−0.86)	(−0.86)	(−0.62)	(−0.62)
Location	0.283	0.093	0.223	0.072
	(1.15)	(1.16)	(0.90)	(0.90)
Cons	−2.597 ***		−4.133 ***
	(−3.80)		(−3.09)
Robustness Check
Wald chi2	78.68		70.19
Prob > chi2	0.00		0.00
Log likelihood	−217.99		−212.03
Pseudo R2	0.18		0.20
Goodness fit	73.70%		76.07%
Multicolinearity	33.79		41.93
N	422		422

* p < 0.1, ** p < 0.05, *** p < 0.01. Z value in parentheses. 30 < Coldiag2 < 100 means multicollinearity which has no effect on the results.

displays the model findings and the results of checks for robustness.

The goodness-of-fit tests of Model 1 and Model 2 were 73.70% and 76.07%, respectively. Multicollinearity tests showed that there existed a certain degree of multicollinearity in Models 1 and 2 but had no effects on the results. To further ensure the robustness of Model 1 and Model 2, we deleted the control variables based on Model 1 and Model 2 (Appendix

Table A1. Results of the Probit model evaluation and marginal effects without control variables.

Variables	Model 3		Model 4
	Probit	mfx	Probit	mfx
dum_age ($\le$35)	−0.661 ***	−0.199 ***	−0.505	−0.155
	(−2.64)	(−3.34)	(−0.75)	(−0.89)
dum_age ($age\ge$ 60)	0.153	0.055	3.054 **	0.856 ***
	(1.00)	(0.99)	(2.03)	(5.45)
Social capital	0.0976 **	0.035 **	0.274 *	0.096 **
	(2.19)	(2.21)	(1.92)	(2.00)
Social capital * dum_age ($\le$35)	-	-	−0.307 **	−0.107 **
	-	-	(−2.03)	(−2.12)
Social capital * dum_age (age$\ge$60)	-	-	−0.0166	−0.006
	-	-	(−0.25)	(−0.25)
Cons	−1.367 **		−3.099 **
	(−3.12)		(−2.16)

* p < 0.1, ** p < 0.05, *** p < 0.01.

). In addition, we also performed a stepwise regression (Appendix

Table A2. Results of the Probit model evaluation and marginal effects without insignificant variables.

Variables	Coeff	Z Value
dum_age ($\le$35)	−0.687 *	(−2.53)
dum_age ($age\ge$ 60)	2.703	(1.89)
Social capital	0.254 *	(1.96)
Social capital * $dum\_age$ ($\le$35)	−0.291 *	(−2.03)
Chinese level	0.342 **	(3.13)
Ratio of non-livestock income	−1.002 ***	(−3.31)
Livestock	0.181 **	(3.25)
Hayfield	−0.170 ***	(−5.35)
Cons	−3.931 **	(−3.02)

* p < 0.1, ** p < 0.05, *** p < 0.01.

). The results were consistent with Model 1 and Model 2, confirming the accuracy of the regression results.

In both Model 1 and Model 2, social capital had a significant impact on encouraging households to rent grassland, confirming H1. The marginal effect of social capital on land rental market participation probability is 0.028, implying that a one-unit increase in social capital leads to a 2.8% improvement in participation probability. Model 2 included an interaction term for social capital and an age dummy variable, and social capital was less helpful for the youngers to rent land than for their middle-aged counterparts. This confirms H2. In addition, the age dummy variable showed that compared with the middle-aged group, the propensity for the youngers to rent land resources was lower, which was consistent with empirical fact, while it was easier for herders older than 60 to rent land.

A higher Chinese level signifies stronger bargaining power in the land rental market, which helps to reduce transaction costs and establish a land lease. The flock size was shown to have a positive and significant influence on all samples’ rental decisions. This is also intuitively predicted because greater flock sizes necessitate more fodder, thereby increasing the likelihood for the herder to rent grassland. The area of contracted hayfields significantly affected all herders’ rental decisions, while contracted pastures did not. A larger hayfield produced more forage, thus resulting in a lower likelihood of renting grassland. The policy factors had no significant effect on land rental choices.

6. Discussion

To address our variable of interest, we first discuss social capital. Most existing studies mainly pay attention to the advantages of social capital [19,20,42]. Few studies have focused on different segments of the population benefited by social capital and how. For example, Bezu and Holden [3] found that the youngers in rural Ethiopia have limited access to agricultural land because of land market restrictions. Chamberlin and Ricker-Gilbert [43] argued that transaction costs discourage land rental behaviors, but social capital can help rent land by reducing transaction costs [44].

This study confirmed these findings. Social capital improved herders’ involvement in the grassland rental market, as expected by H1. Moreover, we further found that social capital had a weaker effect on the youngers’ rental decisions compared with their middle-aged counterparts, confirming H2. This was explained by two reasons. One reason is that the youngers might not possess enough social capital to allow them to rent land as easily as middle-aged herders did. As stated by a 30-year-old herder: “Now it is quite difficult to rent grassland locally. With the recent rapid rise in beef and mutton prices, many herders cannot rent land, which is especially true for us young herders. Most of the youngers who come back to Gacha (village in Inner Mongolia) are well educated, but they have been away from Gacha for a long time. They don’t have many acquaintances in Gacha and don’t have the channels or information to rent land.” The second reason is that younger people often build their social capital through interpersonal networks with peers [40]. As a result, the elder generations tended to rent lands to their own peers rather than to the youngers. This is also confirmed by our field survey: “It is difficult for the young to rent land. Most of the pasture is owned by elderly people, and they are not willing to rent the grassland to the youngers. This is mainly because the young have less prestige in Gacha, and the older adults are wary of renting their pastures to the young. They also worry that the young will abuse their rented pastures.”

Young adults in pastoral areas generally face the problem of insufficient social capital and intergenerational gaps. How to help young adults improve their social capital to alleviate the plight of land inaccessibility has become the key to helping them better integrate into animal husbandry production. One way is to pay more attention to returning young herders and provide more opportunities to increase their social capital. As some village chiefs in Inner Mongolia said, “Several college students come back to graze in Gacha every year. Although we don’t want our children to come back to graze and bear hardships, we still welcome them if they want to come back. After all, grazing is very hard. Older people will develop health conditions after grazing for a long time, and they still need more help from young people. Going out for training and other opportunities are basically reserved for Gacha’s college students.” If Gacha can provide more opportunities for returning young herders, it will help them quickly improve their popularity and prestige, thus contributing to the rapid accumulation of their social capital.

Another way is to promote grassland co-management or cooperative groups among young herders to help them gain grassland. We interviewed a 35-year-old herder named Wu in the Urat middle banner, who is a leader of a cooperative group; he stated: “I studied and worked in Japan for several years. Later, due to customs, income and other reasons, I chose to return to the pastoral area and become a herder in 2014. Because of my good education, I have my own ideas on animal husbandry operations and am willing to adopt new technologies. In addition, I have also learned relevant knowledge about modern animal husbandry locally and applied it to production practice. In 2015, my income from animal husbandry was approximately 500,000 yuan; by 2020, it reached 900,000 yuan. As the scale of animal husbandry expanded, I wanted to set up a cooperative to concentrate the animal husbandry resources in the surrounding areas and lead other herders to make money.” Because of Wu’s excellent animal husbandry operations and the considerable income it brings, the herders in the surrounding area gradually believed in Wu, and this trust laid a good foundation for him to establish the cooperative. In May 2020, Wu and five young herders jointly established a cooperative. The cooperative has a total grassland area of 55,000 mu and raises 3000 sheep, 80 cattle, and 18 horses. In 2020, the income of the cooperative was nearly CNY 2 million, and the average net income of each household was approximately CNY 150,000. “Since 2014, more than 30 young herders have returned from cities to Gacha, and they include members of the cooperative,” Wu said. He continued, “The success of the cooperative is largely because members are all young people and like-minded; the cooperative helps the youngers get access to grassland so they can engage in animal husbandry.”

7. Concluding Remarks

In this paper, we applied a probit model to a dataset with 422 sampled households from Inner Mongolia and found that social capital significantly improved the possibility of herders renting land. The youngers’ social capital is significantly weaker in helping them rent land compared with the social capital of the middle-aged group. In China, and in other pastoral areas of the world, however, the youngers are faced with the relatively difficult situation of accessing land. Given land scarcity and land market restrictions [3], lack of social capital is attributed as a major reason for this. In the long term, with increasing populations and more or less stable land area, the contradiction between people and land will be more prominent. In this case, better social capital can not only help young people obtain access to land resources but also allow them to explore other opportunities. For example, grassland co-management can allow the same grassland area to produce higher outputs and to save costs on the labor force. Furthermore, better social capital may help herders link their livestock production with markets through cooperation, bringing higher profit. The increased output value means that herders can generate the same income with less land.

Although this study was conducted in Inner Mongolia, we believe that the observations and conclusions derived from the large survey data may provide important insights into land policy interventions ensuring the livelihoods of the youngers and the future development of similar areas. The main limitation of this paper is that the data used are relatively old. We will supplement and improve them in future research and add samples from different pastoral areas.