How Do Cooperatives Alleviate Poverty of Farmers? Evidence from Rural China

Abstract

Farmers’ cooperatives play an important role in enabling small farmers to integrate into modern agriculture. Based on the survey data of 7200 farmers in four provinces of China, this paper uses the multi-dimensional poverty measurement method and the instrumental variable method to reveal the transmission mechanism and multi-dimensional poverty reduction effect of farmers’ cooperatives in deep poverty-stricken areas in China to realize joint agricultural empowerment through the supply chain. The results show that farmers’ cooperatives play an important role in enhancing small farmers’ financing, technology application, market sales, and rights decision-making. Every increase in the degree of interest connection between the two will help farmers reduce the multi-dimensional poverty level by 12.3%, and the mitigation effects on material poverty, ability poverty, and rights poverty are between 10% and 13%. Compared with agricultural cooperatives with weak organizational service capacity, cooperatives with high organizational service capacity have a more significant multi-dimensional poverty reduction effect on farmers, which leads to farmers obtaining financing and improving their ability and income. In addition, the difference in benefit coupling between agricultural cooperatives and farmers of different leading bodies also results in the heterogeneity of multi-dimensional poverty alleviation effects of farmers. To a certain extent, the supply chain is affected by the risk endowment of members, the supply chain’s poverty alleviation performance is affected by members’ supply chain’s poverty alleviation performance, and there is a certain risk transfer and “distribution failure”. Therefore, optimizing the benefit coupling structure and risk sharing mechanism between farmers’ cooperatives and farmers has become an important way to break the pattern of “the strong are always strong” and “distribution failure” for supply chain members and farmers.

1. Introduction

In developing countries, agriculture is the primary subsistence source for most rural low-income families [1]. To effectively lift farmers out of poverty and raise their income, agricultural cooperatives must continuously optimize their governance during the development process. To address the issue of farmers’ inability to meet their material, ability, and rights needs, agricultural cooperatives and farmers can effectively regulate the rights and obligations of stakeholders by establishing a reasonable and stable benefit coupling. Spreading and distributing the monetary and non-monetary benefits of resource utilization between the national economy and its related stakeholders is conducive to broad-based inclusive growth. This helps to form a community of interests, mobilize all stakeholders’ enthusiasm, and improve farmers’ current poverty situation, which is caused by various factors [2].

The services provided by cooperatives are gradually diversifying in the digital age. Taking the market as a guide and optimizing the allocation of resources, farmers’ cooperatives that adapt to the size of industrial operation are thriving, and they have formed a beneficial coupling of risk and sharing and win–win cooperation with farmers. Dhiab et al. [3] found that it is essential for cooperatives to give consulting services to smallholders, which is beneficial to rural development innovation and knowledge production. It is also considered a reaction to global agriculture’s numerous issues [4]. However, there is a common problem: the main contributions of supply chain organizations in the implementation process are often ignored, such as the mediation in farmers’ decision-making, service diversification, and differentiated development of agriculture [5]. On the one hand, cooperatives vary greatly in terms of motivation, resource endorsement, service capacity/function, leadership, and the relationship between core and ordinary members. By the end of December 2021, the number of officially registered farmers’ cooperatives in China reached 2.219 million, comprising and aiding nearly half of all farmers. Alongside the large number of excellent or genuine cooperatives falling into the definition of the International Cooperative Alliance (ICA), unsurprisingly, there are also many “shell cooperatives” (or faked, existing in name only) caused by many complicated factors.

On the other hand, the full impact of the outbreak of the COVID-19 pandemic in developing nations has yet to be determined. It remains difficult to anticipate its size and nature. This applies not just to infections, illness, and mortality, but also to the downstream effects on sectors such as agriculture and agricultural trade, including the effects on smallholders, commercialized agriculture, and agricultural value chains [6]. In addition, market risk, financial risk, and technical risk will significantly impact the material demand and income of smallholders.

Given the nature of poverty traps, including numerous dimensions or causes of rural poverty [7], alleviating poverty through cooperative development necessitates a diversified but intertwined strategy to free smallholders from multiple restraints. Accordingly, this paper draws attention to the bonds between smallholders and cooperatives in poor areas, referring to interconnections and benefit coupling between them in order to better use resources and opportunities, both internally and externally, provided by either government agencies or other stakeholders. The financial services provided by agricultural supply chain finance, for instance, can effectively promote the organic connection among small-scale farmers, modern agricultural development, and the large-scale operation of agricultural operating entities [8] and overcome the defects of agricultural cooperatives due to their small scale and insufficient ability to resist risks [9]. Furthermore, with the intermediary of agricultural cooperatives between smallholders and financial institutions, smallholders’ credit availability and access can be significantly improved [10]. However, there is also the possibility that core members of cooperatives may squeeze or occupy the financial credit resources of the agricultural supply chain [11], resulting in farmers’ needs not being effectively met. Therefore, technical and sale services carried out by agricultural cooperatives can be used as an effective means for poverty alleviation [12,13].

Having recognized the importance of cooperative development for sustainable rural development and poverty alleviation, the Chinese government has issued a series of favorable policies, including financial subsidies, to support the development of cooperatives since 2007. This has resulted in the rapid growth of registered cooperatives in China, reaching 2.2 million in total by the end of 2020, or nine times over that in 2009. Bearing in mind the unevenness of cooperative development by geographic region, agricultural sector, and local economic development across the county, the Chinese government launched a national campaign, namely “targeted poverty alleviation” (TPA), to eradicate rural poverty nationwide by the end of 2020. This involves a large scale of mobilization and coordination for all types of resources (technology, information, investment, finance, and personnel) from all sectors (public, private, and non-government organizations) across different geographic boundaries (from the coastal to the middle and west provinces) to support rural poor households and villages in 832 government recognized poverty-stricken counties in West China. Among many objectives or indicators (e.g., infrastructure, housing, education, health, food and water security) assigned by the central government for successful poverty alleviation projects, cooperative development has been listed as an important indicator to ensure that the rural poor have access to or can gain benefits from the public services and various kinds of support provided by the government or poverty alleviation agents. In other words, among several poverty alleviation strategies under the umbrella of TGA, poverty alleviation through cooperative development is an important means to help poor rural households to gain access to finance, technology, and the sale of their agricultural products.

Nearly half of the world’s smallholders (about 230 million) are currently living in China, which provides many research samples for exploring the bonds or benefit coupling between farmers and cooperatives. Cooperative development plays a vital role in farmer empowerment, agricultural modernization, rural transformation, and revitalization in China. Despite the decisive government intervention with and large-scale public participation in poverty alleviation, China still has a long way to go in terms of cooperative development, farmers’ organizations, and transformation. Based on a large-scale questionnaire survey involving 7200 farmers in 18 state-designated impoverished counties of three provinces (Sichuan, Yunnan and Jiangxi Provinces) and Chongqing City in China in 2015 and 2016, this study examines the benefit coupling mechanism between smallholders and agricultural cooperatives and its impact on poverty reduction in terms of financing, agro-technology, sales, and participation in decision-making.

This paper differs from previous relevant literature studies. It focuses on analyzing the interest connection between farmers’ cooperatives and farmers. It compares and analyzes whether differences in the degree of interest connection between farmers and farmers’ cooperatives in the Chinese context will have a different impact on their poverty relief effectiveness, using the international poverty standard classification as a reference for farmers with different interest connection degrees. Relevant research can better provide targeted support measures for farmers who are members of farmers’ cooperatives of different poverty types and improve farmers’ credit availability and welfare level, especially in different types of poverty—material poverty, ability poverty, and right poverty—where there is greater heterogeneity in their effects. This paper in particular verifies the alleviation effect of cooperative interest connection on farmers’ poverty, investigates heterogeneity and mechanism, and broadens the connotation and application scope of existing poverty theory from the perspective of interest connection.

2. Theoretical Background

Sen developed a multi-dimensional poverty theory, breaking through the limitation of traditional theory, using income poverty as the sole indicator of poverty. The theory proposes the concept of ability poverty, which states that poverty is defined as a lack of functional welfare. Preference policies, subsidy policies, and special assistance policies may result in a lack of functional welfare. Subsidies for land circulation, tax relief, cooperative loan subsidies, subsidies for purchasing agricultural machinery, and special industrial support projects will affect member farmers’ functional welfare differences. In other words, personal welfare is guaranteed by ability, while poverty is caused by lack of ability [14]. Therefore, poverty is the economic situation represented by income, and multi-dimensional poverty transcends economic aspects.

According to the stakeholder theory developed in the 1960s in Western nations and used in select corporate governance models, stakeholders are those individuals and groups who make unique investments and assume certain risks in the production activities of firms [15]. Their activities have the potential to influence or change corporate goals, and the process of achieving those goals also influences them. In this paper, farmers in rural China rely on rural agricultural cooperatives to achieve poverty reduction goals, while agricultural cooperatives rely on farmers’ input and participation to survive, thereby realizing the overall interests of stakeholders. Various forms of connections among stakeholders reflect income distribution and rights relations [16]. Therefore, based on the multi-dimensional poverty theory and stakeholder theory, this paper divides the benefit coupling mechanism into four aspects—financing connection, agro-technology connection, sales connection, and decision-making connection (see Figure 1)—and explores the multi-dimensional poverty mitigation by agricultural cooperatives.

2.1. Financing Connection and Material Poverty

According to Leroy and Pop [17], financing connection refers to the two-way interaction between the financial system and the economy. The financial system’s role is critical for any economy because capital is allocated to economic agents with productive investment prospects [18]. However, commercial banks and other financial institutions hardly want to enter the rural financial market due to the high economic risks they have to bear, which causes the economy in rural areas to lag. In the process of the development of China’s rural agriculture, on the one hand, the smallholders have to bear natural disasters and market risks because they have limited assets to mortgage [18]. On the other hand, the bank credit policy has failed to adapt to the demand for capital by the smallholders because the banks find it difficult to measure the repayment ability and credit conditions of smallholders at a low cost [19]. In recent years, with the rapid development of financial technology and the continuous business model innovation of online finance, the farming community has been able to help rural farmers to increase the possibility of applying for full credit through the supply chain of agriculture financial credit or by providing financing services for its members. The members of agricultural cooperatives may enjoy a guarantee, negotiate with agricultural enterprises, and provide input for agriculture [20].

2.2. Agro-Technology Connection and Material Poverty

As Abebaw and Haile [21] mentioned, the agro-technology connection is short for agricultural technology. It includes agricultural intensification, which means adopting advanced means of cultivation to increase agricultural production and provide farmers with a favorable price [22]. Knowledge and experience are two obstacles to the development of small-scale farmers [23]. The rapid advancement of agricultural modernization leads to an increase in demand for agro-technology, which promotes the establishment of agro-technology connections between agricultural cooperatives and smallholders. Agricultural cooperatives can provide technical support to make up for the deficiency of relevant stakeholders in technical knowledge and to reduce their learning and trial-and-error costs [24], especially for rural young people who are caught in the dilemma of having no intention or ability to engage in agricultural production. Young people also lack farming ability and technical support, so farming is not the best choice for young people in rural China; they would rather go out to work than stay in rural areas. The central government of China has repeatedly issued policies to promote training in agricultural science and technology, the cultivation of new types of professional farmers, and the increase the investment in developing agro-technology training [25].

In addition, agricultural cooperatives are conducive to implementing policy projects to benefit farmers. Integrating and expressing farmers’ scientific and technological needs on the platform of agricultural cooperatives can effectively match the government’s supply with smallholder needs. The membership of agricultural cooperatives increases the use of chemical fertilizers. It improves the possibility of adopting high-quality seeds [21,26], indicating that establishing an agro-technology connection in agricultural cooperatives can affect farmers’ awareness of the independent application of agro-technology and form the concept of modern agricultural development. In agricultural machinery technology, the local government organizes centralized technical training for smallholders by introducing the agro-technology connection mechanism of agricultural cooperatives to reduce high-technology input costs because of the small-scale operation and insufficient ability to apply modern technologies [27,28]. At the same time, agricultural cooperatives accumulate human capital by increasing investment in agricultural technology education services. The improvement of rural human capital is beneficial to poor areas to improve the efficiency of food production technology, thus speeding up poverty alleviation, promoting farmers’ endogenous development ability, and guaranteeing the organic connection between farmers and modern agricultural technology [23]. The radiating and driving effect of agricultural cooperatives on surrounding farmers and rural development can undoubtedly promote employment, realize income increases for rural farmers, and alleviate poverty [29]. Regarding information technology, with the rapid development of the agricultural Internet of Things and big data, researchers have begun to pay attention to the impact of e-commerce on rural poverty. Agricultural cooperatives can actively strive for policy support and supply chain finance credit, develop the rural e-commerce poverty alleviation service platform, and provide farmers with e-commerce training and professional information to alleviate material poverty [30].

2.3. Sales Connection and Material Poverty

Most farms in China are small and labor-intensive units that lack market competitiveness [31]. Due to the long distance and poor sales channels, the agricultural industry chain has a serious problem of information asymmetry [32]. There is a lack of smooth information exchange channels and trust mechanisms between the producers and sellers. On the one hand, with the increase of purchasing power, consumers are paying more attention to food quality and safety to easily purchase affordable high-quality agricultural products. On the other hand, farmers are trying to find stable sales channels. Agricultural cooperatives are shortening supply chains by engaging in vertical integration [11]. They are usually some of the best platforms to increase the opportunities for small farmers to enter the market. By connecting small farmers with markets, agricultural marketing cooperatives play an important role in assuring farmers of competitive producer prices and helping poor farmers to benefit more from the spillover effects of agricultural cooperatives. First, agricultural cooperatives can improve farmers’ market connection ability. Farmers are linked through agricultural cooperatives to solve farmers’ competitive disadvantages in bargaining [33,34]. They aim to expand their profit margin, so they usually buy high-quality inputs at lower prices and sell high-quality agricultural products at higher prices [35]. Second, agricultural cooperatives control the quality of agricultural products, thus realizing the traceability of agricultural products and ensuring food safety [36] and can gradually build a green brand of agricultural products [37]. Farmers can become good business partners of market actors downstream of the agricultural supply chain [38] to establish long-term stable contact with the consumption terminal, promote the circulation of agricultural products, and reduce the risk of unsalable agricultural products. Finally, with the adoption of advanced agro-technology, the production capacity of farmers can be improved. It is good for farmers to engage in agricultural production to be self-sufficient and increase their income.

2.4. Participation in Decision-Making and Rights Poverty

The right to vote is a basic right for cooperative members to exercise democratic management. Improving the decision-making mechanism is key to promoting the standardized operation and sustainable development of rural agricultural cooperatives. The internal governance structure of agricultural cooperatives generally consists of three main bodies: the member’s congress, the board of directors, and the board of supervisors. The agricultural cooperative usually adopts the one-person, one-vote system when making decisions on important matters [39]. In other words, each member has one vote. In the actual decision-making process, due to the lack of a perfect system and mechanism of agricultural cooperatives, democracy is a mere formality in rural China. The decision-making power is concentrated in the hands of only a few people represented by the chairperson of the agricultural cooperative, and the decision-making of one dominant group replaces democratic decision-making. Although this organization can improve the decision-making efficiency to some extent and make up for the defect of smallholder consciousness, it is still difficult to mobilize members’ enthusiasm to participate in the agricultural cooperatives. On the other hand, the participation degree and discourse rights of farmers are not high, and the problem of allowing ordinary members to free-ride cannot be solved, so it is hard to fundamentally change the current situation of peasant households. The participation of peasant households directly affects the understanding and identification of decision-making and implementation efficiency. Whether the will of peasant households is fully expressed through the member’s (representative) congress and whether their enthusiasm is mobilized are two important problems to focus on when investigating the democracy of decision-making in the agricultural cooperatives. They are also important to measure the poverty degree of peasant households’ rights.

3. Previous Studies and Research Hypotheses

Cooperatives can help smallholders in many ways. Because most farmers in China still rely on traditional self-sufficient production, some of them find it difficult to access external finance, which has led to poverty [40]. Agricultural cooperatives can provide financing services for their members and help farmers to obtain financial credit for the agricultural supply chain [41]. Then, financial resources can be reallocated from rich urban centers to poor rural areas, which can broaden farmers’ financing channels and positively impact rural living standards [42], reducing poverty, increasing social welfare, and alleviating urban–rural inequality [43]. Because of the poor infrastructure in the countryside and the weak technological capability of farmers, who might find it difficult to cope with modern technology and to change consumer demand [24], it is important to promote the agro-technology connection between farmers and agricultural cooperatives by providing agricultural technology training. Wu et al. [44] and Ding et al. [25] found that the application of agricultural technology played a significant role in improving rural household income and alleviating rural poverty based on a survey of farmers in Yunnan Province of China. The agro-technology connection can effectively improve farmers’ agricultural skills, transfer knowledge, encourage farmers to adopt advanced technologies, and help them to improve their production efficiency and labor income [45]. At the same time, the weak ability of members to sell their agricultural products without the support of agricultural cooperatives has become an important factor affecting farmers’ material poverty in rural China. By improving farmers’ bargaining power in the market and concentrating funds and resources through cooperative enterprises, the sales connection between farmers and agricultural cooperatives can reduce marketing costs and realize economic benefits for farmers [46,47]. It can also enable smallholders to overcome the powerful oligarchic investment companies to charge lower prices and effectively solve the problem of information asymmetry [48]. Cooperatives can market agricultural products more efficiently and increase farmers’ production and income, alleviating farmers’ material and ability poverty [49]. Ajates [50] believes that agricultural cooperatives are more moral and value-oriented than private enterprises, attaching more importance to the sustainability of cooperation between farmers and cooperatives. In the new type of agricultural operation, such as cooperatives, decisions on important issues need to be made by voting by members. This is helpful to make farmers participate in collective action and improve their rights to speak more actively. Based on the previous studies, we propose the following hypothesis:

Hypothesis 1 (H1).

The higher the degree of benefit coupling between smallholders and agricultural cooperatives, the more conducive this is to improving farmers’ multi-dimensional poverty alleviation degree.

Among the four dimensions of benefit coupling, the closer the financing connection is, the more conducive it is to improve the credit availability of smallholders and solve the problem of low income caused by insufficient lending. The closer the agro-technology connection, the more likely it is to improve agro-technology for smallholders. The closer the sales connection, the more efficient it will be to market agricultural products. Combining these three connections strengthens the effect of reducing smallholders’ material and ability poverty. Since the decision-making connection is closely related to smallholders’ rights poverty, the democratization of the “one person, one vote” approach may be more conducive to improving smallholders’ status in the agricultural cooperative and alleviating their rights poverty.

With the gradual development of agricultural cooperatives, benefit coupling has become an effective way to avoid risks and reduce losses. The impact of benefit coupling on material poverty and ability poverty is reflected in the following aspects. In terms of financing, information constraints, small transaction volume, high transaction cost, low capital, and insufficient credit are all unfavorable for individual smallholders to profit from the market alone. Meanwhile, smallholders in developing countries face many credit constraints and imperfect markets, which affect their investment decisions [12]. Agricultural cooperatives can use their credibility and diversified financing channels to increase their collateral value [51], help smallholders to enhance their ability to repay and reduce credit defaults, improve their credit availability, and break through financing constraints brought by loan ceilings [52]. They can also help to solve the problem of low income and the inability to receive higher education and medical services because of insufficient funds allocated to farmers by the government [53]. In terms of agro-technology, farmers’ scientific cognition of new technology risk is different from their self-cognition [15]. Although some smallholders do not adopt the new agricultural technology directly on their own, spillover effects do exist [54]. This means that when other farmers in the same village use the new agricultural technology, the farmers who do not use it can also benefit through extensive communication and contact in the rural communities. Gao et al. [55] also point out that technical training enhances the exchange of information among smallholders in the same village and promotes adopting new agricultural technology among untrained farmers. Smallholders who actively participate in training activities to learn new agricultural technology can use high-tech tools to form a positive development trend in crop cultivation. This is conducive to improving the labor efficiency of farmers, increasing the scale of cultivated land, reducing the employment cost of technology personnel, and bringing more benefits to the smallholders.

In terms of sales, agricultural cooperatives hold a large market share in the distribution of agricultural products from farms to end consumers. They provide smallholders with a unified and efficient sales channel and give them a better position in price negotiations and access to the markets they cannot enter alone [56]. At the same time, agricultural cooperatives can train the sales capability of farmers so that farmers can cope with the fluctuating prices of agricultural products on the market [57]. On the other hand, Qian [58] finds that relative equity requirements play a key role in decision-making. When smallholders’ relative equity investment is lower than a certain threshold, they will participate in the agricultural cooperative, which shows that cooperative organizations attach great importance to fair demand when making decisions.

Hypothesis 2 (H2).

Among the four dimensions of benefit coupling, financing, agro-technology, and sales connection are relatively more conducive to reducing the material poverty and ability poverty of farmers. Decision-making connections may enhance the additional voting rights of capable persons and alleviate rights poverty.

Under the influence of the development background and establishment time, different forms of agricultural cooperatives differ in their organizational service ability to a certain extent due to the adequacy of public funds and the balance of interests [59]. The organizational service capacity of farmers’ cooperatives is reflected in their ability to connect government and social resources, effectively connect members and farmers, and improve the service efficiency of cooperatives, especially in playing a bridge role between small farmers and modern agricultural markets. In the connection of interests between smallholders and agricultural cooperatives, the agricultural cooperatives with weak organizational service capacity will inevitably accept all agricultural products from their members and farmers, resulting in limited supply and damaging agricultural economic environment cooperatives. On the one hand, agricultural cooperatives with higher organizational service ability can better manage the cooperatives’ assets and provide more financial support to smallholders. On the other hand, they can improve the professional skills of smallholders in an organized and directional way to facilitate different agricultural population groups to get better services [60]. Therefore, compared to those with weak organizational service ability, the agricultural cooperatives with strong organizational service ability are more able to lead farmers to obtain financing and improve their ability and income, thus reducing multi-dimensional poverty more quickly.

At the same time, due to the heterogeneity among the agricultural cooperatives with different leading subjects, the effect of the benefit coupling mechanism on the multi-dimensional poverty reduction of peasant households is also heterogeneous in China. The capable people lead the first type of agricultural cooperatives. They can effectively lead farmers to use advanced scientific planting methods and agricultural technologies to improve agricultural output, expand the scale of farming with their rich experience in scientific planting, and reduce rural poverty with agricultural technology [14]. The second type of agricultural cooperative is run by village cadres, which can contribute to alleviating rural poverty mainly by enhancing the democratization of decision-making. Village cadres, who usually have high public trust in villages, can effectively drive smallholders to participate in the decision-making of agricultural cooperatives. Besides, the leader of the agricultural cooperative, who is usually a member of village committees, can better represent the wishes of smallholders [61]. Therefore, we put forward the following hypothesis:

Hypothesis 3 (H3).

There is a heterogeneity of organizational service ability in alleviating multi-dimensional poverty through the benefit coupling mechanism of agricultural cooperatives.

4. Methodology

4.1. Data

The data in this paper come from the survey conducted by the research group on rural financial poverty alleviation in state-designated impoverished counties in China (including Sichuan Province, Chongqing City, Jiangxi Province, and Yunnan Province) in 2015 and 2016. By investigating the changes in the livelihoods of officially registered poverty-stricken households, we examined the characteristics of various types of poverty-stricken areas in different provinces, such as the concentrated and contiguous poverty-stricken areas, remote ethnic poverty-stricken areas, old revolutionary base areas, ethnic minority poverty-stricken population gathering areas, and ethnic poverty-stricken population gathering areas. These are the core places of deep poverty-stricken areas in China and are highly representative for analyzing the poverty alleviation effect of agricultural supply chain financial credit.

First, a questionnaire survey was conducted with farmers who had obtained or were willing to participate in farmers’ cooperatives’ supply chain financial credit in the sample national poverty-stricken counties. The farmers’ questionnaire covered their family characteristics, income, annual agricultural production output, loan demand, and credit availability. Second, according to the local farmers’ cooperatives offering the agricultural supply chain financial credit, the questionnaire survey also included questions on the scale of cooperatives, assets status, profitability, financial support projects and subsidy policies, various incentive schemes, industrial structure, and the credit supply of formal financial institutions. Third, we conducted semi-structured interviews and questionnaires with the local commercial banks/microfinance institutions, local governments’ poverty alleviation offices, and financial management departments engaged in the agricultural supply chain financial credit business. Ten farmers’ cooperatives were selected for investigation in every county. A total of 180 farmers’ cooperatives were investigated in the four provinces. The 40 farmers were randomly sampled according to the local members who had participated in or were willing to participate in the financial credit of the agricultural supply chain. Finally, in selecting sample villages, we focused on the cooperatives in the poverty-stricken areas where local financial institutions had carried out the agricultural supply chain finance activities relatively well.

A total of 7200 questionnaires were distributed in the survey. In total, 6790 questionnaires were valid after eliminating those missing key information, with a recovery rate of 94.32%. Of these, 5003 farmers had participated in or were willing to participate in the agricultural supply chain financial credit, accounting for 73.67% of the total valid samples, while 4611 farmers had repaid their loans on time, accounting for 92.16% of the total number of farmers participating in the financial credit of the agricultural supply chain. In the valid data, the maximum size of a cooperative formed from the 6790 cooperative members was 169, and the minimum was 64. On average, each cooperative had 88 members. We also collected the data on GDP per capita and rural poverty incidence in four provincial areas from the local statistical bureaus.

4.2. Measurement

4.2.1. Multi-Dimensional Poverty Comprehensive Index of Farmers

In the aspect of the decomposition and measurement of the multi-dimensional poverty composite index of the smallholders, this paper uses the A-F multi-dimensional poverty measurement method proposed by Alkire and Foster [62], which is based on several axiomatic standards. The design of the index also refers to the article of Zhang et al. [63]. The detailed measurement steps are as follows.

First, we determined the measurement dimensions. A multi-dimensional poverty comprehensive index of smallholders was developed based on the three dimensions of material poverty, ability poverty, and rights poverty. Eight measurement indicators were selected, holding one-eighth weight each: net income per capita, cultivated land area per capita, food expenditure, health status, years of education of farmers, ownership concentration, the decision-making power of important matters of agricultural cooperatives, and the village cadre’s status. The construction of the multi-dimensional poverty index in the paper adopts substitution weighting.

The second step was to set and identify the deprivation threshold of each dimension. We measured the multi-dimensional poverty composite index of $n$ farmers in $t$ dimensions $\left(t=8\right)$, used $p_{im}$ to express the value of the $i\mathrm{th}$ farmers $\left(i\in R,i=1,2,\dots ,n\right)$ in dimension $m$ $\left(m\in R,m=1\dots \dots , 8\right)$, and used $v_m$ as the critical value of deprivation. If $p_{im}$ did not reach the critical value of $v_m$, $p_{im}=1$ (farmers were identified as poor in this dimension); otherwise, it was 0.

The total deprivation score $s_i$ was obtained by weighting the deprivation value $p_{im}$ of the $i\mathrm{th}$ farmer in $t$ dimensions, that is, $s_i={{\displaystyle \sum }}_{m=1}^t{p}_{im}\times w_m,s_i\in \left[0,1\right]$. We set the poverty dimension critical value $c$, and if the $i\mathrm{th}$ farmer was identified as in poverty in $c$ dimensions or more (that is, the frequency of $p_{im}=1$ was greater than or equal to $c$), then the $i\mathrm{th}$ farmer was listed as being in multi-dimensional poverty. $R\left(c\right)$ represented the incidence of poverty, $I\left(c\right)$ represented the intensity of poverty, and $mpi$ represented the multi-dimensional poverty composite index of farmers. The calculation formula of $mpi$ was as follows:

$$mpi=R\left(c\right)\times I\left(c\right)$$

(1)

$$R\left(c\right)=\frac{{{\displaystyle \sum }}_{i=1}^n{q}_{im}\left(c\right)}{n}$$

(2)

$$I\left(c\right)=\frac{{{\displaystyle \sum }}_{i=1}^n{s}_i\left(c\right)}{{{\displaystyle \sum }}_{i=1}^n{q}_{im}\left(c\right)\times t}$$

(3)

where $q_{im}\left(c\right)$ means that the farmer is defined as multi-dimensional poverty, and $s_i\left(c\right)$ is the total score of the $i\mathrm{th}$ farmers’ deprivation. When $s_i\left(c\right)\ne 0$, $q_{im}\left(c\right)=1$; otherwise, $q_{im}\left(c\right)=0$. From Formula (1), we can see that the multi-dimensional poverty index of farmers is the product of the poverty incidence and intensity.

4.2.2. Model Construction

First, in order to estimate the comprehensive effect of the benefit coupling of agricultural cooperatives on the multi-dimensional poverty alleviation of smallholders in rural China, we applied the same weight to the connection of financing, agro-technology, sales, and decision-making, and then calculated the benefit coupling index of agricultural cooperatives. The OLS regression model was as follows:

$$mpi={\alpha }_0+{\alpha }_{1}ll{i}_i+{\alpha }_2{V}_i+{\delta }_i$$

(4)

where $mpi$ is the multi-dimensional poverty composite index of farmers, $ll{i}_i$ is the benefit coupling index of the $i\mathrm{th}$ farmer, $V_i$ is the related control variable of the $i\mathrm{th}$ sample, and ${\delta }_i$ is the random disturbance term.

Second, to measure the impact of the various dimensions of benefit coupling between agricultural cooperatives and smallholders on poverty alleviation, we preliminarily analyzed the effectiveness of the various dimensions of benefit coupling on reducing the multi-dimensional poverty of farmers. The OLS regression model is as follows:

$$mpi={\beta }_0+{\beta }_{1}f{l}_i+{\beta }_{2}t{l}_i+{\beta }_{3}s{l}_i+{\beta }_{4}d{l}_i+{\beta }_5{Z}_i+{\sigma }_i$$

(5)

where $f{l}_i$ indicates whether the $i\mathrm{th}$ farmer gets the agricultural supply chain financial credit and the financing services provided by agricultural cooperatives, which measures the effect of financing connection on alleviating the smallholders’ multi-dimensional poverty. $t{l}_i$ measures whether the $i\mathrm{th}$ farmer receives the agro-technology training provided by the agricultural cooperatives, which reflects the impact of agro-technology connection on farmers’ multi-dimensional poverty. $s{l}_i$ measures the ability of the $i\mathrm{th}$ farmer to sell agricultural products by themselves when separating from the agricultural cooperative, which reflects the effect of sales connection on the multi-dimensional poverty comprehensive index of farmers. $d{l}_i$ refers to the voting method of the $i\mathrm{th}$ farmer to participate in cooperative agricultural affairs, which measures the degree of alleviating the farmers’ multi-dimensional poverty by decision-making connection. $Z_i$ refers to the relevant control variables of the $i\mathrm{th}$ sample. ${\sigma }_i$ is a random disturbance term.

4.3. Variables Description

4.3.1. Dependent Variable

In order to reveal the poverty situation of smallholders in rural China, this paper deconstructed the multi-dimensional poverty index of farmers from the aspects of material poverty, ability poverty, and rights poverty by using the OLS model. The specific measurement method of the index is as mentioned in

Table 1. Multidimensional poverty index of rural households and the explanation of deprivation threshold.

Dimension	Index	Determination and Explanation of Deprivation Threshold	Weight	Incidence of Poverty
Material poverty	Net income per capita	Net income per capita < RMB 2300 is 1; otherwise, it is 0	1/8	83.95%
	Cultivated land area per capita	Cultivated land area per capita < 1.43 mu is 1; otherwise, it is 0	1/8	60.39%
	Food expenditure	If proportion exceeds 40% of the total expenditure, it is 1; otherwise, it is 0	1/8	54.36%
Ability poverty	Health	Good = 0; disease = 1	1/8	46.97%
	Years of Education	senior high school or above = 0; junior high school or below = 1	1/8	34.27%
Rights poverty	Ownership concentration	Equity concentration ≥ 30% is 1; otherwise, it is 0	1/8	88.10%
	Right of cooperatives’ important matters	President decision = 1; membership meeting or Council decision = 0	1/8	46.27%
	Village cadre or not	No = 1; yes = 0	1/8	42.87%

.

4.3.2. Independent Variables

Considering the benefit coupling between agricultural cooperatives and smallholders in financing, agro-technology, sales, and decision-making as independent variables, we gave the same weight to obtain the benefit coupling index of agricultural cooperatives. Among them, the financing connection included whether the farmers obtained the agricultural supply chain financial credit and whether the cooperatives provided financing services.

Table 2. Descriptive statistics.

Variable Name		Variable Description	Mean	Variance	Max	Min
Dependent variable
Multi-dimensional poverty index		See Table 1	0.248	0.305	1.488	0.024
Independent variables
Financing connection	Whether farmers get credit	Yes = 1; No = 0	0.303	0.459	1	0
	Whether cooperatives provide financing	Yes = 1; No = 0	0.302	0.459	1	0
Agro-technology connection	Whether cooperatives provided training	Yes = 1; No = 0	0.186	0.389	1	0
Sales connection	The ability of the smallholder to sell agricultural products without cooperatives	General and above = 0; 1 otherwise	0.577	0.494	1	0
Decision-making connection	The voting method of agricultural cooperatives	One person, one vote = 1; One share, one vote = 0	0.391	0.488	1	0
Control variables
Cooperative level	Scale	Total number of members	105.827	20.679	160	26
	Fixed assets	Fixed assets of cooperative	36.268	8.961	56	8
	Profitability of cooperative	from 1 (very poor) to 5 (very strong)	3.110	0.360	5	2
Farmer level	Age	Age of farmer	46.868	9.174	70	22
	Gender	male = 1; female = 0	0.274	0.513	1	0
	Whether farmers have financing history	Yes = 1; No = 0	0.484	0.500	1	0
Region level	GDP	Regional per capita GDP	4704.78	2022.46	8820.31	2764.86
	Incidence of rural poverty	Rural poverty/total rural population	6.113	3.537	12.7	2.0

shows that 30.3% of farmers had agricultural supply chain financial credit, and 30.2% of agricultural cooperatives provided financing services to the smallholders. The agricultural technology connection mainly focused on whether the agricultural cooperatives provided agricultural technology training for the smallholders. Only 18.6% of smallholders received the agro-technology training provided by the agricultural cooperatives. In terms of sales connection, Table 2 indicates that the ability of the smallholders to sell agricultural products by themselves without joining agricultural cooperatives was generally weak. The decision-making connection mainly focused on the voting method of agricultural cooperatives. Only 39.1% of cooperatives used the “one person, one vote” method.

4.3.3. Control Variables

In order to prevent the influence of other related variables on the independent variables and reduce the influence on the dependent variable, the control variables were selected from the three levels of agricultural cooperatives, farmers, and regions. The control variables of agricultural cooperative characteristics included the total number of members, fixed assets, and profitability. Farmer’s characteristics included age, gender, and financing history. The regional characteristics covered GDP per capita and the incidence of rural poverty. Generally speaking, the higher the number of cooperative members, the more fixed assets, the stronger the profitability, and the higher ability cooperatives have to provide financing, sales, and other services for the farmers. The longer the financing history of farmers, the lower their multi-dimensional poverty composite index. The older the farmers, the higher level of the multi-dimensional poverty composite index. The higher the GDP per capita, the lower the incidence of rural poverty and the better the situation of poverty alleviation in these areas.

4.4. Endogeneity Discussion

In this part, we try to put forward a preliminary solution to the endogeneity problem that may occur in the model so as to ensure the robustness of the regression results in the following pages. On the one hand, the omitted variables may be related to the benefit coupling mechanism of the agricultural cooperatives and may affect the multi-dimensional poverty index of rural households simultaneously, resulting in the missing variable bias. On the other hand, the model might have a reverse causality relationship, that is, the differences in material, ability, and rights poverty might lead to different levels of farmers’ credit availability and their ability to sell agricultural products without the support of the agricultural cooperatives. Generally speaking, the allocation and the degree of risk of farmers’ assets are important references for credit institutions to consider whether to provide loans to farmers. Farmers with a high degree of poverty may be blocked from the threshold of financial credit due to high credit risks, which is especially obvious among farmers with prominent material poverty. For farmers with higher degrees of ability and rights poverty (those who are in poor health, accept less education, or have no rights in cooperatives), it is very likely that there exist natural disadvantages during the selling of their agricultural products, such as a lack of efficient sales channels, no chance to come into contact with larger markets, inability to accept large sales order, etc., leading to a low sales ability of agricultural products. In order to overcome the above negative effect on the unbiasedness, consistency, and validity of the OLS regression coefficients, we plan to adopt an instrumental variable (IV) to solve the potential endogeneity problem of the model to obtain more reliable and robust empirical results.

For model (5), which estimates the retarding effect of the association of various dimensions on the multi-dimensional poverty index of farmers, this paper used the degree of brand awareness of cooperatives as an instrumental variable. The values from 1 to 4 represent “very low”, “low”, “average”, and “high” levels of cooperative brand awareness for the model estimation. The agricultural cooperatives with relatively high brand reputations have more resources to support their member farmers to sell their agricultural products and improve the sales connection. Since the brand awareness of cooperatives has no direct relationship with the multi-dimensional poverty index of farmers, it meets the exogenous requirements.

5. Results

5.1. Basic Regression Analysis

In this study, we first analyzed the overall effect of the index of benefit coupling of agricultural cooperatives on the comprehensive index of farmers’ multi-dimensional poverty in rural China. In

Table 3. The multidimensional poverty reduction effect of benefit coupling: OLS regression.

Variable	Comprehensive Index of Farmers’ Multi-Dimensional Poverty		Material Poverty			Ability Poverty			Rights Poverty
	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)	(10)
Index of benefit coupling of agricultural cooperatives	−0.123 *** (0.013)	−0.121 *** (0.014)	—	—	—	—	—	—	—	—
Financing connection	—	—	−0.117 ** (0.008)	−0.116 ** (0.011)	−0.124 *** (0.011)	−0.124 *** (0.011)	−0.130 *** (0.150)	−0.126 *** (0.159)	—	—
Agro-technology connection	—	—	−0.133 *** (0.010)	−0.132 *** (0.010)	−0.137 *** (0.008)	−0.137 *** (0.008)	−0.109 * (0.013)	−0.107 * (0.013)	—	—
Sales connection	—	—	−0.109 * (0.005)	−0.109 * (0.005)	−0.112 *** (0.016)	−0.112 *** (0.016)	−0.133 *** (0.150)	−0.132 *** (0.150)	—	—
Decision-making connection	—	—	−0.111 ** (0.008)	−0.011 ** (0.008)	—	—	—	—	−0.146 *** (0.010)	−0.142 *** (0.011)
Characteristic control variable	no	yes	no	yes	no	yes	no	yes	no	yes
Constant term	0.263 *** (0.013)	0.237 *** (0.027)	0.179 *** (0.011)	0.186 *** (0.022)	0.934 *** (0.089)	0.934 *** (0.089)	1.134 ** (0.007)	0.905 ** (0.027)	0.427 *** (0.030)	0.427 *** (0.030)
Pseudo R2	0.002	0.002	0.032	0.041	0.009	0.008	0.000	0.026	0.003	0.007
N	6790	6718	6790	6718	6790	6718	6790	6718	6790	6718

Note: Standard errors are in parentheses. ***, ** and * indicate significance at the level of 1%, 5% and 10%, respectively.

, column (1) reports the regression results without control variables. Column (2) reports the regression results with control variables. The results show that each unit increase of the index of benefit coupling reduces the farmers’ multi-dimensional poverty composite index by 2.7%, which is significant at 1%. Column (2) shows that there is no significant difference between column (1) and column (2) after adding the control variables, which indicates that the benefit coupling has a significant effect on the multi-dimensional poverty alleviation of farmers in rural China, which is consistent with Hypothesis 1.

Second, we demonstrated the effect of financing, agro-technology, sales, and decision-making in the benefit coupling of agricultural cooperatives on alleviating farmers’ multi-dimensional poverty, as indicated in column (3) and column (4). The results indicate that compared to the farmers who did not form financing connections with agricultural cooperatives, the multi-dimensional poverty composite index of farmers who developed a financing connection with the agricultural cooperatives was reduced by 1.7%. This was significant at the level of 5%, with other factors unchanged. The conclusion is consistent with Mojo et al. [24]. The multi-dimensional poverty index of the farmers who had the agro-technology connection with the agricultural cooperatives was 3.3% lower than that of the farmers who did not have the connection. This was stable and significant at the level of 1%. In terms of the decision-making connection, the empirical results show that the farmers who had a decision-making connection with the agricultural cooperatives had a mitigation effect of 1.1% on the multi-dimensional poverty composite index, which was significant at the level of 5%.

Third, we estimated the poverty alleviation effect of each benefit coupling mechanism on the material poverty, ability poverty, and rights poverty of farmers. The material poverty index, ability poverty index, and rights poverty index of farmers were calculated by weighing the corresponding indexes in Table 1, and the empirical results are reported in columns (5)–(10) in Table 3. Among the four dimensions, the financing, agro-technology, and sales connections have mitigation effects of 2.4%, 3.7%, and 1.2% on the farmers’ material poverty, which are significant at the level of 1%. They have mitigation effects of 3.0%, 0.9%, and 3.3% on the farmers’ ability poverty, respectively. The decision-making connection has a mitigation effect of 4.6% on alleviating farmers’ rights poverty. The conclusion is consistent with Ma et al. [27]. Therefore, Hypothesis 1 and Hypothesis 2 are confirmed.

Since the p values of the above regression results were generally small, we reasonably suspected that it was caused by the endogeneity problem in the model. Therefore, we used the instrumental variable method, as we mentioned in Section 4.4, to analyze the endogeneity of the model.

Table 4. The multidimensional poverty reduction effect of benefit coupling: endogeneity test.

Variable	Multi-Dimensional Poverty Index of Farmers
	(1)	(2)	(3)
Financing connection	−0.017 ** (0.008)	−0.015 ** (0.008)	−0.015 ** (0.008)
Agro-technology connection	−0.033 *** (0.010)	−0.032 *** (0.008)	−0.032 *** (0.008)
Sales connection	−0.009 * (0.005)	−0.008 * (0.007)	−0.008 * (0.007)
Decision-making connection	−0.011 ** (0.008)	−0.010 ** (0.006)	−0.010 ** (0.006)
Control variables	yes	yes	yes
Regional control	yes	yes	yes
Constant term	0.263 *** (0.013)	0.261 *** (0.024)	0.261 *** (0.024)
n	6790	6790	6790
Pseudo R2	0.002	—	—
F value	2.09	—	—
Wald chi2(9)	—	24.00	24.00
p-value	0.022	0.008	0.008

Note: Standard errors are in parentheses. ***, ** and * indicate significance at the level of 1%, 5% and 10%, respectively.

shows the results of the endogeneity test by using the degree of brand awareness of agricultural cooperatives as an IV. Column (1) reports the OLS regression results, while column (2) reports the regression results by using the two-stage least square method (2SLS). After introducing the IV, the model passes the Hausman test. Thus, the null hypothesis that all explanatory variables are exogenous can be rejected at the significance level of 10%. Members’ ability to break away from the agricultural cooperative and sell agricultural products by themselves, which represents the sales connection, is considered an endogenous variable. The empirical results in column (2) are robust after excluding the endogeneity effect. After excluding the influence, the retarding effect of the association of various dimensions on the multi-dimensional poverty index of farmers is still significant and negative, which is consistent with the results in column (1). Column (3) reports the empirical results of using the limited information maximum likelihood (LIML) method, which is less sensitive to weak instrumental variables. It remains consistent with the results in column (2), which confirms the hypothesis of no weak instrumental variables. In terms of effect, the order is as follows: agricultural technology linkage > financial linkage > decision-making power linkage > sales linkage. The possible reason for this is that as an economic cooperation organization, farmers’ cooperatives can often obtain technical guidance from the government’s agricultural technology extension personnel, become a link to connect members of farmers’ technology promotion and application, and can better drive members to accept new technologies to improve the quality and efficiency of agricultural products. At the same time, farmers operate through cooperatives.

5.2. Mitigation Mechanisms

The above analysis of empirical results has shown that the benefit coupling mechanism of agricultural cooperatives has a significant and negative impact on the multi-dimensional poverty of farmers in rural China. In order to further clarify the mitigation process of farmers’ material, ability, and rights poverty, we also added farmers’ credit availability. Whether the cooperative hires technical personnel, a fixed marketing channel of agricultural products, or additional voting rights for capable persons respectively illustrate the mitigation mechanisms of financing connection, agro-technology connection, sales connection, and decision-making connection for reducing the farmers’ multi-dimensional poverty.

5.2.1. Mitigation Mechanisms of Material and Ability Poverty

A financing connection improves the credit availability of smallholders in rural China. Column (1) in

Table 5. Mitigation mechanisms of financing, agro-technology, and sales connection on material and ability poverty of farmers.

Explained Variable	Farmers’ Credit Availability		Whether Cooperative Hires Technical Personnel		Fixed Marketing Channel of Agricultural Products
	(1)	(2)	(3)	(4)	(5)	(6)
Financing connection	0.161 *** (0.913)	0.168 *** (0.920)	—	—	—	—
Agro-technology connection	—	—	−0.034 ** (0.015)	−0.034 ** (0.015)	—	—
Sales connection	—	—	—	—	0.272 *** (0.093)	0.276 *** (1.657)
Control variables	yes	yes	yes	yes	yes	yes
Regional control	yes	yes	yes	yes	yes	yes
Constant term	0.796 *** (1.906)	0.867 *** (1.002)	0.362 *** (0.037)	0.279 *** (0.116)	1.725 (0.552)	1.789 (0.560)
Pseudo R2	0.001	0.001	0.008	0.003	0.003	0.004
n	6790	6790	6790	6790	6790	6790

Note: Standard errors are in parentheses. ***, ** and * indicate significance at the level of 1%, 5% and 10%, respectively.

shows the OLS regression results, and column (2) indicates the 2SLS regression results after excluding the endogeneity effect. As we can see from Table 5, the credit availability of farmers increased significantly by 16.1% at the 1% level for each additional unit of financing connection. The farmers who have a financing connection can rely on the agricultural cooperatives’ high credibility and scale of capital to form stable and safe financing channels, raise sufficient agricultural production, and reduce their own agricultural production funds, as well as reducing their credit risk. Eventually, they can reach the goal of solving the material poverty problem due to the lack of start-up funds and addressing the ability poverty issue due to the lack of individual resource endowment.

The agro-technology connection reduces the human capital expenditure of agricultural cooperatives. For each additional unit of agro-technology connection, the probability of cooperatives employing technical personnel decreased significantly by 3.4% at the level of 5%. The agro-technology training the cooperatives provided for their member farmers can help to efficiently allocate human resources [22]. The human capital expenditure can be reduced through mutual agro-technology assistance within the cooperatives. The agricultural cooperatives can invest more funds into their daily agricultural production and operation to minimize cost and maximize income. As a result, the cooperatives and farmers can increase their production and income, which makes it possible for the farmers to receive better education and medical services.

The sales connection offers marketing channels of agricultural products by the cooperatives to the local farmers. The proportion of fixed sales channels of agricultural products increased significantly by 27.2% at the level of 1% for each unit increase in sales connection. The formation of sales connections makes it possible for agricultural cooperatives to give full play to their advantages in information and resources [11]. The agricultural cooperatives can help link the farmers to the market, establish the marketing channels for their agricultural products, reduce the risk of unsold agricultural products, and eventually bring about a significant income increase for the farmers [64]. Column (6) shows that the regression coefficient is still significant, which verifies the interpretation of the empirical results in column (5). This indicates that the benefit coupling mechanism of agricultural cooperatives can indeed promote the increase of farmers’ income in various ways to alleviate their material and ability poverty.

5.2.2. Mitigation Mechanisms of Rights Poverty

In the same way, a binary variable—“additional voting rights for capable persons”—was added to explore the mechanism of decision-making connection represented by the agricultural cooperative counting method on alleviating farmers’ rights poverty. The empirical results in

Table 6. Mitigation mechanisms of decision-making connection on rights poverty of farmers.

Explained Variable	Additional Voting Rights for Capable Persons
	(1)	(2)
Decision-making connection	0.031 *** (0.010)	0.029 *** (0.010)
Control variables	yes	yes
Regional control	yes	yes
Constant term	0.115 *** (0.032)	0.166 ** (0.069)
Pseudo R2	0.008	0.008
n	6790	6790

Note: Standard errors are in parentheses. ***, ** and * indicate significance at the level of 1%, 5% and 10%, respectively.

indicate that compared to the voting by share, the probability of capable persons gaining additional voting rights increased by 3.1% and was significant at the level of 1% when the agricultural cooperative adopted the one-person, one-vote counting method. Because of the existence of decision-making connections, the decision-making of agricultural cooperatives has become more democratic. The additional voting rights obtained by capable persons representing the ordinary farmers can significantly avoid the power concentration problem of agricultural cooperatives. Every farmer can enjoy equal decision-making rights. Column (2) presents the regression results with the IV. The regression coefficient is still significant. This indicates that the mitigation mechanism of the benefit coupling mechanism of agricultural cooperatives to reduce the rights poverty of farmers does exist in rural China by increasing the additional voting rights of capable persons.

6. Heterogeneity Analysis

6.1. Heterogeneity of Organizational Service Capability

In this study, we further explored the heterogeneous impact of the benefit coupling mechanism of agricultural cooperatives on the multi-dimensional poverty alleviation of farmers in terms of different organizational capability agricultural cooperatives possess.

Table 7. Multidimensional poverty reduction effect of benefit coupling: heterogeneity of organizational service capability.

Variable	Multi-Dimensional Poverty Index of Farmers
	Very Poor	Relatively Poor	General	Good	Very Good
Financing connection	0.054 (0.295)	−0.023 * (0.154)	−0.035 * (0.213)	−0.402 ** (0.303)	−0.524 *** (0.036)
Agro-technology connection	−0.027 (0.291)	−0.037 ** (0.014)	−0.043 ** (0.021)	−0.058 ** (0.023)	−0.061 ** (0.03)
Sales connection	0.004 (0.035)	0.010 (0.010)	−0.011 ** (0.021)	−0.026 ** (0.018)	−0.031 ** (0.024)
Decision-making connection	−0.017 (0.035)	−0.050 (0.011)	−0.113 * (0.017)	−0.212 * (0.019)	−0.329 ** (0.025)
Control variables	yes	yes	yes	yes	yes
N	289	3308	1317	1156	720
R2	0.016	0.004	0.006	0.006	0.016
p value	0.873	0.146	0.581	0.657	0.174

Note: Standard errors are in parentheses. ***, ** and * indicate significance at the level of 1%, 5% and 10%, respectively.

shows the multi-dimensional poverty reduction effect of the benefit coupling mechanism of agricultural cooperatives according to the heterogeneous service capability of sample agricultural cooperatives, which were classified into five categories: “very poor”, “relatively poor”, “general”, “good”, and “very good”. As Table 7 indicates, compared to the agricultural cooperatives with poor organizational service capability, those with higher organizational service capability have a more significant multi-dimensional poverty reduction effect on the farmers. The higher the organizational service capability that agricultural cooperatives possess, the more efficient the asset management will be. This is consistent with Deng’s [31] research findings that professional cooperatives perform better in technology demonstration and standardization. This means that they could greatly mobilize farmers’ enthusiasm to participate in the financial credit and cultivate their financial literacy. Therefore, agricultural supply chain finance and credit can be used more effectively to improve the living standards of poor farmers in rural China. On the other hand, agricultural cooperatives with higher organizational service capability can provide better agro-technology training for the local farmers. The multi-dimensional poverty alleviation effect is more obvious because the agricultural cooperatives with higher organizational service capability can guide farmers to master and apply agro-technology to the farming process more efficiently. Moreover, they can help farmers to make better decisions about their production and operation to achieve a win–win situation.

6.2. Heterogeneity of Leading Entities

The agricultural cooperatives with different leading entities may alleviate the farmers’ poverty at different levels of benefit coupling. According to the different leading entities of cooperatives, they can be classified into three main types: large farmers, village cadres, and enterprises. As

Table 8. Multidimensional poverty reduction effect of benefit coupling: heterogeneity of leading subjects.

Variable	Multi-Dimensional Poverty Index of Farmers
Leading Subjects	Capable People	Village Cadres	Enterprises
Financing connection	−0.032 * (0.026)	−0.005 ** (0.017)	−0.006 *** (0.018)
Agro-technology connection	−0.067 *** (0.024)	−0.013 * (0.015)	−0.036 ** (0.015)
Sales connection	−0.001 ** (0.011)	−0.004 ** (0.007)	−0.003 *** (0.007)
Decision-making connection	−0.033 ** (0.019)	−0.004 *** (0.012)	−0.005 (0.012)
Control variables	yes	yes	yes
n	979	2498	2845
R2	0.007	0.002	0.003
p-value	0.023	0.624	0.629

Note: Standard errors are in parentheses. ***, ** and * indicate significance at the level of 1%, 5% and 10%, respectively.

shows, the agricultural cooperatives led by large farmers mainly alleviate farmers’ poverty by forming agricultural technology connections. Large farmers generally have a certain scale of planting and harvesting. Their planting and breeding technology is more scientific to help increase the production and income of smallholders to alleviate their poverty by disseminating the advanced agricultural technology. Second, the village cadres have a good mass base, which is conducive to connecting with the farmers and mobilizing their enthusiasm to participate in agricultural cooperatives. Because they can represent the wishes of ordinary farmers more, they can improve the democratic decision-making of cooperatives. Third, the agricultural cooperatives led by the enterprises have demonstrated a stronger poverty alleviation effect in terms of developing financing connections and sales connections because they can form a certain scale effect by financing the smallholders’ production orders. They usually possess very strong market competitiveness and can easily obtain bank financial credit. Moreover, their marketing and distribution system can deal with the market risks more effectively, thereby contributing to the poverty reduction of farmers in rural China. This is consistent with the findings by Brugere et al. [2].

7. Discussions

The benefit coupling mechanism between agricultural cooperatives and smallholders plays an important role in promoting the effective connection between smallholders and the market, contributing to alleviating the multi-dimensional poverty in rural China. It has great significance for developing long-term multi-dimensional poverty alleviation mechanisms and fostering new types of agricultural business entities to increase the incomes of smallholders effectively. According to the empirical results, the main research questions raised in the paper can be addressed as follows. First, despite the deficiencies in developing agricultural cooperatives, they play a key role in getting rid of multi-dimensional poverty and increasing smallholders’ income in rural China. Second, based on different mitigation mechanisms, the benefit coupling between agricultural cooperatives and smallholders in financing, agro-technology, sales, and decision-making can form a joint force in the agricultural production process and the internal operation management system of cooperatives and jointly bring the cooperatives into full play. Finally, the operation of agricultural cooperatives has demonstrated heterogeneous poverty reduction effects in rural China because of the differences in their organizational service capability and leading entities. The cooperatives that provide better comprehensive organizational services usually perform better and bring about a higher multi-dimensional poverty reduction effect.

Based on the empirical results, we put forward the following recommendations for the sustainable development of agricultural cooperatives in rural China. On the one hand, the governments should establish and improve the evaluation system of agricultural cooperatives, inspect their operation regularly, and implement targeted measures for fake cooperatives to reduce the abuse of government subsidies and public funding. On the other hand, it is important to help smallholders to connect with modern agricultural technology and develop effective benefit coupling with agricultural cooperatives to promote the coordinated development of agricultural supply chain finance. In terms of financing connections, the government should formulate effective policies to encourage cooperatives with high organizational service capability and credibility to provide better financing services for the farmers. In terms of agro-technology connection, the government should guide the agricultural cooperatives to disseminate the relevant agro-technology knowledge and optimize the learning curve of farmers. In terms of sales connection, the government should support the agricultural cooperatives to open up sales channels for agricultural products, attract market funds actively, and guide agricultural products to go global. For participation in the decision-making process, a key indicator for the maturity of cooperative development, the government can guide the agricultural cooperatives to promote the proper separation between voting rights and management rights so that the farmers can have a better awareness of their voting rights as cooperative members. Their voting rights should be encouraged and better protected.

8. Conclusions

This study investigates the significance of benefit coupling between agricultural cooperatives and smallholder farmers in relieving multi-dimensional poverty and bringing rural China toward shared prosperity. This article examines the poverty reduction benefits of benefit coupling between agricultural cooperatives and smallholder farmers in terms of financing, agro-technology, sales, and decision-making, using survey data from four provinces in China from 2015 and 2016.

First, the benefit coupling of financing, agro-technology, sales, and decision-making between agricultural cooperatives and smallholder farmers can reduce the multi-dimensional poverty of rural farmers. This helps to reduce the credit risk and enhance the credibility of agricultural cooperatives to improve the credit availability for smallholder farmers. Then, it solves the problem that the farmers have labor but are short of capital. Besides, the agricultural cooperatives can make up for farmers’ lack of experience in agricultural cultivation and management, reduce trial and error costs so that the farmers can better resist natural risks, and reduce losses. On the other hand, agricultural cooperatives can help the farmers cultivate high-quality varieties, increase output, and expand the scale of farming. In addition, the sale of agricultural products by the farmers themselves can gradually improve their management ability and reduce the economic losses caused by the decay of agricultural products. To a certain extent, such a sales connection can help to alleviate the multi-dimensional poverty of farmers themselves. The “one person, one vote” counting approach can help the farmers to improve their status in the agricultural cooperatives, helping to reduce the farmers’ competitive disadvantage and playing a role in alleviating their multi-dimensional poverty.

Second, the beneficial coupling of financing, agro-technology, sales, and decision-making aids in reducing material poverty and poverty of ability and rights. The agricultural cooperatives have built a benefit coupling mechanism with farmers in financing, agro-technology, and sales so that farmers’ production and operation activities have financial security, technical security, and output security. The farmers can carry out production and operation activities more smoothly, which is conducive to increasing their production and income, expanding arable land, and reducing their food expenditure. In addition, the “one person, one vote” approach for the farmers to participate in the decision-making regarding the internal affairs of agricultural cooperatives means that the farmers’ voting rights can be further affirmed, which can help them to decide on the important matters of agricultural cooperatives and create opportunities for them to become the village cadres.

Finally, compared to agricultural cooperatives with low organizational service ability, those with higher organizational service ability have a more significant multi-dimensional poverty reduction effect on farmers, leading them to obtain financing, improve their ability, and increase their income. Furthermore, different agricultural cooperative leadership entities may alleviate farmer poverty at varying levels of benefit coupling.

This paper focuses on the diverse effects and mitigation mechanisms of agricultural cooperatives in alleviating multi-dimensional poverty among farmers in rural China. Future research could look into how agricultural cooperatives meet the needs of smallholders and how to maximize the effect of poverty reduction even more. We will investigate how to improve the credit rating and brand effect of farmers’ cooperatives, whether different farmers within farmers’ cooperatives will have different poverty reduction effects with the same degree of interest connection, and how to further optimize the interest connection between farmers and cooperatives through the internal governance structure of cooperatives.