How Significant Are the Roles Moral Obligation and Formal Institutions Play in Participatory Irrigation Management?

Abstract

Against the background of the agricultural tax reform and the disintegration of China’s rural collective agriculture system, participatory irrigation management (PIM) is the key to improving irrigation management performance. Based on the survey data of 712 peasant households in the Yellow River basin of China, this study employs multi-group structural equation modeling (SEM) to explore the impact of moral obligation and formal institutions on PIM. The results show that both moral obligation and formal institutions can significantly improve collective action. Collective action can markedly promote irrigation management performance, and the formal institution can significantly enhance the farmers’ moral obligation. Additionally, the results of the multi-group analysis show that the agricultural income level of households and their provinces can regulate the impact of moral obligation and formal institutions on PIM. Therefore, to improve irrigation management performance, strategies of intensifying moral obligation and refining formal institutions are recommended for governments and village committees.

1. Introduction

Northern China is facing severe water shortages, especially in the Yellow River basin. The precipitation of the Yellow River basin in 2020 was 506.9 mm, while the evaporation was between 800 and 1800 mm [1]. The highest levels of evaporation occurred in areas with annual precipitation of less than 400 mm. Therefore, agricultural production in the Yellow River basin is highly dependent on irrigation. To meet agricultural production needs, China has inputted large funds into the construction, management, and maintenance of irrigation facilities. In 2020 alone, this funding reached 818.17 billion yuan (yuan is China’s currency unit; 1 U.S. dollar = 6.62 yuan in May 2022) [2]. Moreover, the strategic plan for rural revitalization (2018–2022) formulated by China clearly states that it is necessary to strengthen the construction of irrigation systems, improve the quality of small-scale irrigation facilities, and promote the long-term sound operation of the projects. However, irrigation facilities still do not meet the real needs of agricultural production, largely because irrigation facilities do not fully perform their functions. Even worse, some of the small-scale irrigation facilities, which serve as the “last kilometer” of irrigation, quickly become “paralyzed” after completion. The main reason for this is that they are not properly operated and maintained [3].

Since the agricultural tax reform and the disintegration of China’s rural collective agriculture system, the responsibility for irrigation management has shifted from rural collectives to individual families [4]. In the unified management mode of the government, the irrigation facilities are mainly constructed and managed by the government. This governance mode has some shortcomings. On the one hand, the government is responsible for the operation and maintenance costs of all irrigation facilities, which has resulted in a serious financial burden on the government. On the other hand, the government formulated the irrigation management scheme without the participation of farmers. This is one of the reasons why the irrigation management scheme cannot effectively meet the actual irrigation demand of farmers. To better meet the irrigation needs of farmers, the government gradually transferred the responsibility of irrigation management to farmers and encouraged farmers to manage irrigation facilities according to their own needs. Thus, participatory irrigation management (PIM) has emerged to replace the unified management mode.

PIM refers to the participation of water users in different aspects of irrigation management, including irrigation system designing, construction, supervision, decision-making, maintenance, evaluation, etc. [5]. This may not only reduce the government’s financial burden on irrigation but also enable the users of irrigation facilities to manage irrigation independently, realizing the sound operation and management of facilities [6]. Under the unified management mode, the government is responsible for the construction, operation, and maintenance of irrigation facilities. However, this does not mean that the government should carry out the specific operation and maintenance work of irrigation facilities. The specific operation and maintenance work is usually undertaken by local farmers. Of course, the government pays or subsidizes these farmers. Despite this, many farmers are still reluctant to participate in irrigation management. One reason is that government payment is too small to create strong incentives. The other reason is that the ownership, operation, and maintenance of irrigation facilities are vested in the government, so farmers feel they have no obligation to participate in irrigation management. As a result, the government does not implement specific operation and maintenance work, the farmers do not participate enough, and the market does not participate, leading to the lack of effective governance of irrigation facilities.

In PIM, farmers are responsible for the operation and maintenance of irrigation facilities. Farmers use irrigation facilities, bear the governance costs, and carry out specific operation and maintenance work, which is conducive to the realization of a virtuous cycle of irrigation facilities being used and managed. This is helpful to improve the quality of irrigation facilities, improve the utilization rate of irrigation water, and ultimately meet the irrigation needs of farmers. Otherwise, it is difficult to comprehensively improve the irrigation management performance at the farm level by relying solely on the government to increase investment in irrigation facilities. However, irrigation facilities are common-pool resources (CPRs) [7]. Their public-good attributes conflict with the private-good attributes of peasant household production, resulting in poor the enthusiasm of farmers to participate in irrigation management [8]. Moreover, the village governance mechanisms are not sound, so they cannot impose strong constraints on farmers’ behaviors toward irrigation management. Consequently, free-riding occurs frequently, which leads to poor PIM performance [9]. Therefore, the key to solving the problem of the management of irrigation facilities lies in the participation of water users [10]. What factors might motivate farmers to participate in irrigation management and improve irrigation management performance? The answer to this question could provide a new way forward for irrigation management and a practical basis for the establishment of a long-term PIM mechanism.

The existing research mainly focuses on analyzing the impact of social capital on PIM. Social capital refers to actors’ social ties, which are characterized by reciprocity and trust, with other individuals or groups [11]. It can restrain individual behaviors through social norms, thereby effectively avoiding free-riding in collective action [12,13]. Similar to social norms, moral obligation can also constrain individual behaviors, but they have different influence mechanisms. Social norms are rules and standards used to guide or restrict the behavior of group members to ensure the consistency of group activities [14]. Social norms affect individual behavior mainly through external pressure, that is, external behavior supervision. Different from social norms, moral obligation is the internalization of good behavior norms, that is, the internal behavior constraint. The research shows that the moral obligation of individuals has a significant impact on their choice for collective action [15], but few scholars study PIM from the perspective of moral obligation.

In addition to moral obligation, institutions are an important factor in the study of irrigation management. Some research focuses on the role of the government in PIM. They argue that the establishment of various farmer cooperation organizations supported by the government can increase the collective action of irrigation management [16]. Besides, government support reduces the pressure on the maintenance cost of farmers’ participation, which can motivate farmers to participate in irrigation management [9]. Nevertheless, the effectiveness of government support depends not only on the policy but also on the implementation of the policy at the village level. Government policy is an important impetus for the formation and improvement of the village institution. A reasonable institution can play a strong role in encouraging and restraining the behaviors of the villagers, to improve the collective governance performance of village public goods [3,17]. Although some scholars have paid attention to the positive role of institutions in collective action, they have not analyzed the formal institution influencing PIM performance. Additionally, research on the relationship between moral obligation and formal institutions has not been thoroughly pursued by scholars.

PIM refers to farmers’ collective participation in different aspects of irrigation management, including the construction, maintenance, supervision, decision-making, and evaluation of irrigation systems [6,18]. The rationality of farmers can lead to the irrationality of the village collective, which results in the collective action dilemma, making it difficult to achieve optimal irrigation management performance. Fortunately, moral obligation and formal institutions provide new ideas for solving the collective action dilemma in PIM. Moral obligation can enhance farmers’ awareness of participating in collective action [19]. The formal institutions can not only regulate farmers’ behavior in PIM through rewards or punishments but also coordinate the relationship between stakeholders [20]. These will ultimately help the achievement of collective action. Therefore, it is of great theoretical and practical significance to study the impact of moral obligation and formal institutions on PIM. Given this, in the context of China’s rural revitalization strategy, based on the perspectives of moral obligation and formal institutions, we conducted an in-depth study on PIM. In rural China, central and local governments directly manage the dendritic system of trunk canals and secondary canals. The village collectives are mainly responsible for tertiary canals and drawing plans for the corresponding water allocation methods and maintenance management. Thus, we mainly focus on the governance of tertiary canals located around rural villages.

The main innovation of this study is the exploration of the influence of moral obligation and formal institutions on PIM, as well as the analysis of the relationship between moral obligation and formal institutions. Based on the institutional analysis and development (IAD) framework, this research first discusses the role of moral obligation and formal institutions in PIM. Secondly, the relationship between moral obligation and formal institutions is analyzed. Thirdly, we employ structural equation modeling (SEM) to analyze the effect of moral obligation and formal institutions on collective action and PIM. Finally, a multi-group analysis was conducted on PIM by taking the level of peasant households’ agricultural income, and the province to which they belong, as the regulating variables.

This article is structured as follows. The next section provides the research hypothesis, while Section 3 provides the theoretical framework. Section 4 describes the data collection process and the variables used in the analysis. Section 5 presents and discusses the empirical results, and Section 6 concludes the paper.

2. Research Hypothesis

2.1. Moral Obligation and Collective Action of Irrigation Management

Morality is the internalization of good behavior, which is the key to achieving the rule of morality in the countryside [21]. Traditional village governance in China mainly relies on morality and folk customs. Usually, villager morality is mainly framed by traditional authorities, including moral models (like Confucius), elite morality, and concomitant shared moral values. Traditional village governance has typical characteristics of self-governance and the rule of virtue [3,22,23]. The current system of village public goods management is mainly determined by the villagers, basically following the rural governance mode of the past. By contrast, modern morality is fragmented [24], leading to an unsatisfactory effect of moral governance on public goods in villages. Therefore, strengthening the moral obligation of farmers is critical to achieving the moral governance of villages. Moral obligation can provide incentives for cooperation between farmers and has a marked impact on the realization of rural governance.

Since PIM is an economic activity, according to the definition of morality in economics by Balan and Knack [25], we define the concept of morality as an aversion to using irrigation systems by free-riding, as well as ethical action toward other human beings, duty, and compliance with moral codes and standards [19]. Research has confirmed the positive role of morality in collective action [26,27,28]. Based on the existing research and the reality of village collective action, the main role of moral obligation in PIM is as follows. Firstly, moral obligation is the sense of morality perceived by individuals when making decisions on whether to implement an action, which is an internalized value [29]. In irrigation management, almost all farmers use irrigation facilities in their agricultural activities. Following the principle of he who uses also maintains, many farmers believe that they have an obligation to participate in irrigation management. Driven by moral obligation, these farmers will participate in collective action rather than free riding. Secondly, morality has a broader meaning which encompasses ethical action toward other human beings, duty, and compliance with moral codes and standards [19]. In PIM, if a farmer does not participate in the maintenance of irrigation facilities, there may be two consequences. One is that irrigation canals cannot effectively transfer irrigation water due to the lack of maintenance, which affects the irrigation water intake of farmers along the irrigation canals. The other is that any maintenance work that is not completed is transferred to other farmers, which undoubtedly brings a burden to others. Both of these situations can have a negative impact on others. To avoid these two situations, ethical farmers will take on responsibility for the good of others, that is, actively participating in irrigation management. Thus, we propose the following hypothesis:

H1: 

Farmers’ moral obligation can promote their participation in the collective action of irrigation management.

2.2. Formal Institution and Collective Action of Irrigation Management

Formal institutions are a series of rules, orders, and formulated codes of conduct [30,31]. In irrigation management, in addition to the Regulations on Farmland Water Conservancy promulgated by the government, there are also irrigation management rules formulated by the village committee or the water user associations (WUA) based on the value orientation and specific beliefs of the local villagers. The formal institution suppresses the possibility of opportunistic behavior, making people’s behavior more predictable [32,33]. Because of the conflict between the public-good nature of irrigation facilities and the private-good nature of peasant household production, the enthusiasm of farmers to participate in irrigation management is low. Under these circumstances, the formal institution plays a significant role in encouraging and restraining farmers’ behavior [34].

In terms of formal institutions, Villager Self-Governance is the basic political institution of rural governance in China. The clearer the Villager Self-Governance institution, the higher the village governance level [3,22,23]. In PIM, the main role of the formal institution is summarized as follows. Firstly, the formal institution can form effective incentives and constraints on farmers by shaping the action situation, thus affecting the decision-making logic of farmers’ collective action [35,36]. Secondly, the formal institution can not only guarantee orderly cooperation and competition among farmers but also make clear the right of each farmer to obtain a consistent income [37,38], thereby motivating farmers’ enthusiasm for collective action. Thirdly, by regulating the relationship between farmers, the formal institution prevents farmers from opportunistic behaviors and ensures the orderly governance of collective action [39]. Hence, we propose the following hypothesis:

H2: 

The formal institution can promote the participation of farmers in the collective action of irrigation management.

2.3. Relationship between Moral Obligation and Formal Institution

The operation and maintenance of irrigation facilities are usually large projects, which can only be completed through the coordinated efforts of the different families of the village [40]. Due to potential opportunistic behavior, irrigation management cannot rely entirely on moral obligation or formal institutions for unilateral governance but needs both moral obligation and formal institutions to play a joint role to manage the situation collaboratively. While the formal institution restrains the opportunistic behavior in irrigation management [3,39], moral obligation can play a supplementary and supporting role from the perspective of the informal institution [15]. In this process, the formal institution may have a positive impact on the moral obligation of farmers. The main reasons are as follows. Firstly, as an important part of the formal institution, the Regulations on Farmland Water Conservancy were implemented on 1 July 2016. Even older than this is the Villager Self-Governance institution, the basic political institution of Chinese villages, which was written into the Constitution in 1982. Moreover, the formal irrigation management institutions formulated by the villages according to their irrigation conditions have a long history. These formal institutions are mandatory and universal [20,41]. They can inhibit the free-riding behavior of farmers in irrigation management so that farmers can show moral behavior. Secondly, these formal institutions have been implemented in rural areas for many years. Farmers’ recognition and compliance with the formal institution will gradually form widely accepted ideas, which will gradually be internalized as moral obligations, and thus have a lasting and stable impact on irrigation management. Therefore, we propose the following hypothesis:

H3: 

The formal institution can positively impact farmers’ moral obligation.

2.4. Collective Actions and Participatory Irrigation Management (PIM) Performance

Farmers’ participation in irrigation management is the collective action that farmers with shared interests or common goals choose to participate in to achieve the successful operation of irrigation facilities [4]. It focuses not only on decision-making but also on process participation, including farmers’ participation in the planning, financing, supervision, management, and maintenance of facilities [42]. In public project management, project behavior refers to the “process” by which actors effectively implement the project as required. Performance is the “outcome” of project implementation [43]. Accordingly, the collective action of PIM is a “process” implemented by the farmers according to the requirements of irrigation management. The performance of PIM is the “outcome”. The “process” certainly affects the “outcome”; in other words, effective collective action could improve PIM performance. Thus, we propose the following hypothesis:

H4: 

The collective action of farmers’ participation in irrigation management can positively impact PIM performance.

3. Theoretical Analysis Framework

3.1. The Institutional Analysis and Development (IAD) Framework

Studies have shown that multiple factors (such as rules, land fragmentation, water shortage, and water intake) have a marked impact on irrigation management performance [10,33,44,45]. The IAD framework is dedicated to explaining the impact of various external factors on the governance of CPRs. It is one of the most representative theories in the field of CPR governance, proposed by Elinor Ostrom and her colleagues as early as the 1980s. It can provide resource users with a set of institutional design schemes and evaluation criteria that promote trust and cooperation [7]. Given this, the IAD framework is particularly suitable for analyzing PIM performance.

The IAD framework consists of external variables, action situations, interactions, outcomes, and evaluation criteria [46] (see Figure 1). The action situation is the core of the IAD framework. It is the social space for interaction between individuals, game cooperation, and the management of CPRs among actors [47]. In the action situation, the actor takes a series of interactive actions, such as resource acquisition, competition and cooperation, conflict, supervision, and negotiation. Ultimately, these interactive behaviors lead to an outcome [48]. Whether the result reaches the expectation of the actor or not needs to be measured by a series of evaluation criteria. At the same time, physical conditions, community attributes, and the rules-in-use are three external variables. They jointly have an important impact on the action situation [47]. The strategic choice of the actor will be influenced by physical conditions, because the physical characteristics of the resource may greatly affect its use [49]. In community attributes, important variables influencing the actors’ action strategy include the norms that are generally accepted by the community and the common understanding of the action structure of members; the heterogeneity of members; the group size; and the allocation strategy of resources [50]. Additionally, farmers mainly participate in irrigation management through investment and labor. Therefore, the household characteristics of farmers, which also fall under community attributes, are also an important factor influencing their participation in irrigation management [9]. Furthermore, the rules-in-use are regulations on what actions are required, prohibited, or permitted [47]. In this research, we focus on how external variables affect PIM performance by influencing collective action.

3.2. Econometric Model

Based on the literature review, this study constructs the SEM of farmers’ participation in irrigation management. As shown in Figure 2, the control variables are observable. Moral obligation, formal institutions, collective action, and PIM performance are latent variables. The formal system is an exogenous variable represented by the symbol ${\xi }_1$. Moral obligation, collective action, and PIM performance are the endogenous variables. Their symbols are ${\eta }_1$, ${\eta }_2$, and ${\eta }_3$, respectively. The relationship between the latent variables is as follows:

$${\eta }_1={\gamma }_{11}{\xi }_1+{\zeta }_1$$

(1)

$${\eta }_2={\gamma }_{21}{\xi }_1+{\beta }_{21}{\eta }_1+{\zeta }_2$$

(2)

$${\eta }_3={\beta }_{32}{\eta }_2+{\zeta }_3$$

(3)

where ${\gamma }_{11}$ and ${\gamma }_{21}$ represent the impact of the formal institution on moral obligation and collective action, respectively; ${\beta }_{21}$ represents the impact of moral obligation on collective action; ${\beta }_{32}$ represents the impact of collective action on PIM performance; and ${\zeta }_1$, ${\zeta }_2$ and ${\zeta }_3$ are residual terms.

4. Data and Variables

4.1. Data Collection

The data are from a field investigation conducted by a research group in the Ningxia, Shanxi, and Shandong provinces in the Yellow River basin from January–March and July–August 2019. The Yellow River basin was chosen as the research area because it is generally dry. The annual precipitation is between 200 and 650 mm, with the precipitation from June to September accounting for about 70% of the precipitation for the whole year. Additionally, the evaporation capacity is very strong, with annual evaporation of around 800~1800 mm. Agricultural production in the Yellow River basin mainly depends on irrigation. The irrigation water mainly comes from the Yellow River, and it flows to the downstream farmland under the force of gravity through concrete-lined or earth canals. In 2020, the agricultural water consumption of the Yellow River basin was 28.17 billion cubic meters, accounting for 64.70% of the total water consumption of the Yellow River [1]. The Ningxia, Shanxi, and Shandong provinces investigated are in the upper, middle, and lower reaches of the Yellow River basin. In 2020, irrigation water consumption from the Yellow River in Ningxia, Shanxi, and Shandong was 3.446, 2.395, and 4.728 billion cubic meters, respectively. These three provinces effectively represent the situation with respect to irrigation management in the Yellow River basin.

The data were collected by combining the methods of questionnaire surveys and villager interviews. Since this study focuses on the PIM at the village level, all data were collected from rural villages. We selected two sample counties in each province based on the statistical information. Subsequently, we obtained the distribution of villages in the sample counties, as well as the details of the management and conditions of irrigation in different villages through the county Water Resources Bureau. According to this information, we eliminated villages without canal irrigation and a participatory irrigation management mode, which guaranteed that our research scenario existed in the sampled villages. After the elimination process, we performed a random sampling method to select 8–10 sample villages from the remaining villages in each county. After arriving at the sampled villages, we obtain detailed information about villages by interviewing the village leaders to eliminate households without involvement in canal irrigation. Then, we employed the random sampling method to select 10–15 peasant households for interviews. We chose the head of the household as the representative of the household to survey because they generally have the ultimate decision-making power in the family. It should be noted that the content of the interview was mainly based on the questionnaire. The questionnaire included queries about physical conditions, group and household characteristics, agricultural income, moral obligation, formal institutions, collective action, and PIM performance. Regarding any questions that household heads did not understand, the investigators gave detailed explanations to them until they could accurately understand and answer the questions. Then, the investigators recorded the answers in the corresponding items. Finally, we investigated 712 rural households in 50 administrative villages as samples.

4.2. Sample Description

The average number of farm households in each village is 431. The average sampled households’ area of irrigated farmland account for approximately 97.88% of their whole cultivated land. The average area of cultivated land is 11.62 mu (mu is a Chinese measure of land area: 1 mu = 1/15 ha). Farmers in the survey mainly grew rice, wheat, and corn. Overall, 98.74% of the surveyed families worked in agriculture, and they had a strong demand for irrigation facilities for agricultural production. Interviewees over the age of 50 accounted for 63.06% of all interviewees, reflecting the aging nature of the rural labor force. The education level was generally low, and 89.89% of the sample had a junior high school education or below. Households with three to five people were the majority, accounting for 60.53%. The information of the samples was consistent with the relevant information of the seventh China Population Census, indicating that the sample of this survey was indeed representative. The basic information on the sample villages and households is shown in Appendix A.

4.3. Variables

The variables in this study mainly included the four latent variables of moral obligation, formal institutions, collective action, and PIM performance. The control variables were observation variables. Some details related to the variables are provided below.

For the moral obligation variable, this study, referring to the research of Sabucedo and colleagues [15], selected the following four items: the sense of obligation, personal satisfaction, autonomy, and objectivity.

Regarding the formal institution, we should note that in irrigation management, the formal institution not only has the function of forming incentives for the farmers but also restrains the farmers’ behavior. It should be noted that what we measured was the villagers’ evaluation of the formal institution, because the formal institution of a village is established, but the villagers’ perception and evaluation of it is the key to affecting their behavior. Referring to the research of Yu and colleagues [51], this study selects four items to measure the formal institution: the degree of information transparency, the clarity of the rights and responsibilities of management and maintenance, the effectiveness of the supervision system, and the effectiveness of the punishment system.

The participation of farmers in irrigation management should not only include their democratic and supervisory roles but also stimulate the knowledge and skills they have acquired to contribute to the success of the collective action. Referring to the research of Fischer and Qaim [39], this study selected three variables to measure the collective action of farmers’ participation in irrigation management: attendance frequency, management and maintenance frequency, and input degree.

PIM performance is directly related to the degree of resolution of the provision problem and the appropriation problem. Thus, we measure the PIM performance from the perspectives of the physical condition of irrigation systems and the capability of water supply. Drawing on the research of Yang and colleagues [9], we selected four variables to measure PIM performance: the condition of the irrigation canal, the degree of timely maintenance, the degree of fair distribution, and the number of irrigation water use disputes.

The selection of control variables in this study was based on the IAD framework. Since rules-in-use are mainly characterized by formal institutions, the control variables mainly included physical conditions and community attributes. The physical conditions were represented by the degree of irrigation water shortage, the severity of the natural conditions of the village, and the distance between the plot and the nearest canal. Community attributes included group characteristics and household characteristics. Group characteristics were measured by the number of farmers sharing the same irrigation canal, the degree of villagers jointly solving irrigation water problems, and the support level of village organizations in irrigation. In irrigation management, the head of the household usually makes participation decisions according to their family situation, so the household characteristics were included in the control variables [9]. They are represented by the family labor shortage and capital shortage. The variables used are specifically defined in

Table 1. Definitions and summary statistics of variables.

Variable	Definition	Mean	Std. Dev	Min	Max
MO1	I feel morally obliged to participate in irrigation management: 1(totally disagree) ~ 5(totally agree)	4.500	0.574	1	5
MO2	I feel satisfied participating in irrigation management: 1(totally disagree) ~ 5(totally agree)	4.216	0.780	1	5
MO3	Whatever others may think, I will participate in irrigation management: 1(totally disagree) ~ 5(totally agree)	4.497	0.569	1	5
MO4	I will participate in irrigation management because I believe this is a positive measure: 1(totally disagree) ~ 5(totally agree)	4.490	0.578	1	5
FI1	Degree of information transparency in irrigation management: 1(very opaque) ~ 5(very transparent)	3.476	0.965	1	5
FI2	Clarity of rights and responsibilities of management and maintenance in irrigation management: 1(not clear) ~ 5(very clear)	3.202	0.943	1	5
FI3	The supervision institution can effectively constrain me to participate in irrigation management: 1(totally ineffective) ~ 5(very effective)	3.206	0.878	1	5
FI4	I would be punished by an institution if I didn’t participate in irrigation management: 1(totally disagree) ~5(very agree)	3.140	0.845	1	5
CA1	Attendance frequency of your participation in meetings related to irrigation management in the village: 1 = never; 2 = once in two to three years; 3 = once a year; 4 = two or three times a year; 5 = more than three times a year	4.049	0.912	1	5
CA2	Maintenance frequency of your participation in irrigation management: 1 = never; 2 = once in two to three years; 3 = once a year; 4 = two or three times a year; 5 = more than three times a year	4.178	0.849	1	5
CA3	The sum of the total labor value and investment level of your participation in irrigation management in the previous year: 1 = not more than 100 yuan; 2 = more than 100 yuan and not more than 200 yuan; 3 = more than 200 yuan and not more than 300 yuan; 4 = more than 300 yuan and not more than 400 yuan; 5 = more than 400 yuan	4.247	0.812	1	5
PIMP1	Condition of irrigation canals: 1(very poor) ~ 5(very good)	3.742	0.867	1	5
PIMP2	The degree of the timely maintenance of irrigation facilities: 1(untimely) ~ 5(timely)	3.402	0.824	1	5
PIMP3	Fairness of irrigation water allocation: 1(not fair at all) ~ 5(very fair)	3.608	0.868	1	5
PIMP4	The number of irrigation water use disputes in the previous year: 1 = more than six times; 2 = five or six times; 3 = three or four times; 4 = once or twice; 5 = never	3.600	0.829	1	5
CV1	Irrigation water is or is not in shortage: 1 = yes; 0 = no	0.560	0.497	0	1
CV2	The natural conditions of the village are or are not harsh: 1 = yes; 0 = no	0.184	0.388	0	1
CV3	The distance between the plot and the nearest canal: 1(very far) ~ 5(very close)	4.833	0.592	1	5
CV4	The number of farmers sharing the same irrigation canal with you: 1 = ten households or less; 2 = eleven to fifteen households; 3 = sixteen to twenty households; 4 = twenty-one to twenty-five households; 5 = more than twenty-five households	3.416	1.179	1	5
CV5	The degree of villagers jointly solving irrigation management problems: 1(never) ~ 5(every time)	3.913	0.943	1	5
CV6	The support level given to you by village organizations in irrigation: 1(not supportive) ~ 5(very supportive)	3.414	0.829	1	5
CV7	The shortage of the agricultural labor force in your family: 1(shortage) ~ 5(abundant)	2.173	1.247	1	5
CV8	The shortage of capital in your family: 1(shortage) ~ 5(abundant)	2.591	1.174	1	5

Note: MO = moral obligation, FI = formal institution, CA = collective action, PIMP= participatory irrigation management performance; CV = control variable. The information shown in this table is all the variables involved in the empirical model.

.

5. Results and Discussion

5.1. Reliability and Validity Analysis

The reliability and validity of the indicators can help to evaluate the scientificity and effectiveness of the evaluation results of the PIM performance model. We tested the indicator system of PIM performance and its influencing factors from two aspects of reliability and validity.

5.1.1. Reliability Analysis

Reliability analyses are used to test the reliability of the index system, which mainly shows the consistency, stability, and reliability of test results. The most used reliability test is Cronbach’s alpha, whose acceptable value should be higher than 0.7. We used SPSS 25.0 software to test the reliability of each latent variable, and the results are shown in

Table 2. The outcome of reliability and validity analysis.

		Moral Obligation	Formal Institution	Collective Action	PIM Performance
Reliability and convergent validity	Cronbach’s alpha	0.877	0.814	0.908	0.824
	CR	0.928	0.881	0.944	0.883
	AVE	0.764	0.652	0.849	0.655
Discriminant validity	Moral obligation	0.874
	Formal institution	0.214	0.807
	Collective action	0.334	0.290	0.921
	PIM performance	0.334	0.274	0.526	0.809

Note: The diagonal represents the square root of the AVE (in bold) while others represent the correlations.

. Cronbach’s alpha value ranged from 0.814 to 0.908, indicating that our indicator system has good reliability.

5.1.2. Convergent Validity Analysis

The convergent validity was applicable to evaluate the internal consistency of the measurement model. It consists of the composite reliability (CR) and the average variance extracted (AVE). When the CR value is higher than 0.6, the index is considered to have good reliability. If the AVE value is greater than 0.5, we believe that the measurement model is sufficiently converged. As shown in Table 2, the minimum CR value and minimum AVE values of moral obligation, formal institutions, collective action, and PIM performance are 0.883 and 0.652, respectively. Both are higher than acceptable levels, indicating that the convergent validity of our measurement model is good.

5.1.3. Discriminant Validity Analysis

Discriminant validity is suitable for judging whether there are differences between latent variables. Fornell [52] believes that the value of the square root of AVE for each latent variable should be greater than the correlation coefficient between the latent variables [52]. The correlation matrix in Table 2 shows that the latent variables have good discriminant validity.

5.2. Overall Structural Equation Modelling (SEM) Analysis

5.2.1. Model Fit

According to the above research, AMOS 25.0 software (IBM, New York, NY, USA) was used to evaluate the goodness-of-fit indices of the SEM model. Figure 3 shows the results of SEM of moral obligation and formal institution affecting PIM performance. The standardized relationships between latent and observed variables are shown in it. The overall goodness-of-fit indices reached an acceptable level (CFI = 0.925, RMSEA = 0.051), indicating that the model and data fit well, and the results can be used to verify the research hypothesis.

5.2.2. Hypothesis Testing

Based on the theoretical model of PIM performance, this study used AMOS 25.0 software (IBM, New York, NY, USA) to obtain the path coefficient among the variables (see

Table 3. Results of the structural model.

Path	Standardized Path Coefficient	t-Value	Inference
H1: Moral Obligation → Collective Action	0.352 ***	9.298	Supported
H2: Formal Institution → Collective Action	0.250 ***	6.274	Supported
H3: Formal Institution → Moral Obligation	0.220 ***	5.236	Supported
H4: Collective Action → PIM Performance	0.600 ***	13.912	Supported
CV1 → PIM Performance	−0.062	−1.729	Not Supported
CV2 → PIM Performance	−0.041	−1.131	Not Supported
CV3 → PIM Performance	−0.018	−0.520	Not Supported
CV4 → PIM Performance	−0.056	−1.621	Not Supported
CV5 → PIM Performance	−0.034	−0.933	Not Supported
CV6 → PIM Performance	0.197 ***	5.348	Supported
CV7 → PIM Performance	0.030	0.863	Not Supported
CV8 → PIM Performance	−0.016	−0.473	Not Supported

*** p < 0.001. Note: In the overall model, RMSEA = 0.051, RMR = 0.04, GFI = 0.925, AGFI = 0.909, NFI = 0.906, CFI = 0.937.

). The standardized regression coefficients of the four hypotheses were all positive and passed the 0.001 significance level test, indicating that the hypotheses were supported.

Among the control variables, the support level given to farmers by village organizations in irrigation can significantly promote PIM performance. A possible reason for this is that the construction and maintenance of irrigation facilities require the investment of plenty of material resources, financial resources, and manpower, which are beyond the bearing range of a single household. The support of village organizations can not only reduce the pressure on farmers’ capital investment but also make farmers feel the importance attached by village organizations to irrigation management [53]. This is helpful to improve the management expectations of farmers and promote the realization of collective action. Furthermore, village organizations can restrain or punish unfair behaviors in PIM, and ensure that farmers can effectively exercise irrigation rights, ultimately improving PIM performance. The possible reason why some of the typical control variables do not pose significant effects on PIM performance might be the style of questioning by our enumerators and, thus, the deliberate responses from our respondents to a certain extent. This might dampen the accuracy and reliability of the research results. Although we trained all the enumerators to ask questions neutrally before the data collection process, this may not have been completely successful.

To better understand the effects (i.e., direct, indirect, and total) of the latent variables in the structural model, the calculated results are summarized in

Table 4. Direct effect, indirect effect, and total effect among latent variables.

Path	Direct Effect	Indirect Effect	Total Effect
Moral Obligation → Collective Action	0.352	--	0.352
Formal Institution → Collective Action	0.250	0.077	0.327
Formal Institution → Moral Obligation	0.220	--	0.220
Collective Action → PIM Performance	0.600	--	0.600
Moral Obligation → PIM Performance	--	0.211	0.211
Formal Institution → PIM Performance	--	0.196	0.196

Note: Total effect = direct effect + indirect effect.

. It shows that the variable with the greatest impact on the collective action is the moral obligation, while the variable with the greatest impact on PIM performance is collective action, followed by moral obligation. A possible reason for this is that the existing formal institutions in rural China are not very reasonable and need to be improved urgently, but farmers are subject to strong moral constraints in the acquaintance society of the village, meaning that their behaviors are greatly affected by morality.

5.3. Multi-Group Analysis

5.3.1. Reasons for Multi-Group Analysis

Farmers’ participation in irrigation management is not only influenced by moral obligation and formal institutions but is also moderated by farmers’ income and regions. On the one hand, the increase in income leads to a strong trend of differentiation, leading to the greatly different behavior preferences of farmers with different income levels, which has an impact on irrigation management [54]. In the income composition, agricultural income has a more obvious impact on the irrigation management behavior of farmers. If the farmers’ income comes from the non-agricultural sector and the village does not force all members to participate in irrigation management, these farmers will most likely not participate in irrigation management. Besides, farmers with different agricultural incomes have different degrees of dependence on irrigation facilities, resulting in different levels of involvement in irrigation management. On the other hand, the participation of farmers in irrigation management varies in different regions. For example, the differences in customs and the degree of the development of WUA in different regions may cause differences in the formal institutions of villages and the moral obligation of farmers, which may have a differentiated impact on PIM performance. Multi-group analysis can not only determine whether the parameters of a measurement model and/or structural model are equivalent across two or more groups but also can be used to determine if samples taken from different sources can be combined into a single data set [55]. Therefore, the multi-group SEM is necessary for exploring how PIM performance varies according to the agricultural income of peasant households and the provinces where they are located.

Based on the literature and experts’ suggestions, and considering the actual situation of farmers in the research area, the agricultural income level of peasant households last year is classified as follows: low agricultural income = 20,000 yuan and below; medium agricultural income = 20,000 to 40,000 (inclusive) yuan; and high agricultural income = more than 40,000 yuan.

Table 5. Results of multi-group analysis.

Path	Agricultural Income
	Low-Income Group (n = 229)	Medium-Income Group (n = 257)	High-Income Group (n = 226)
H1: Moral Obligation → Collective Action	0.347 ***	0.265 ***	0.448 ***
H2: Formal Institution → Collective Action	0.322 ***	0.219 **	0.191 **
H3: Formal Institution → Moral Obligation	0.209 **	0.208**	0.251 ***
H4: Collective Action → PIM Performance	0.729 ***	0.433 ***	0.577 ***
Path	Province
	Ningxia group (n = 240)	Shanxi group (n = 235)	Shandong group (n = 237)
H1: Moral Obligation → Collective Action	0.327 ***	0.367 ***	0.338 ***
H2: Formal Institution → Collective Action	0.436 ***	0.119	0.194 **
H3: Formal Institution → Moral Obligation	0.348 ***	0.236 ***	0.066
H4: Collective Action → PIM Performance	0.537 ***	0.628 ***	0.668 ***

** p < 0.01, *** p < 0.001. “n” represents the number of samples in each group. Note: The influence of the control variables is not shown in this table.

shows the statistics of the agricultural income levels and the provinces.

5.3.2. Multi-Group Test

The multi-group analysis is used to evaluate whether the model adapted to a certain sample is also adapted to other different sample groups. We used a multi-group SEM to test our model for the prediction of PIM performance by agricultural income level and by province. There were six groups: three groups divided by agricultural income from low to high, and three groups of provinces including Ningxia, Shanxi, and Shandong. The parameters of the model are restricted, to obtain the most adaptive path model. This research compares and analyzes the unconstrained model, the measurement weights model, the structural weights model, the structural covariance model, the structural residuals model, and the measurement residuals model. Ultimately, the measurement weights model was selected as the multi-group analysis model in this study. The results are shown in Table 5.

The results of the multi-group analysis are as follows. Firstly, moral obligation significantly promoted collective action in all six groups. Among the three groups of agricultural income, the impact was the largest in the group with a high agricultural income. A possible explanation is that farmers with high agricultural incomes are more dependent on irrigation facilities. Moreover, their financial circumstances enable them to have a strong ability to participate in collective action. Among the three groups of provinces, there was little difference in the degree of impact. A possible explanation is that although the three provinces are in the upper, middle, and lower reaches of the Yellow River basin, villages in China are still an acquaintance society. The behavior of farmers is strongly affected by moral obligation, but this impact does not vary significantly in different regions.

Secondly, the formal institution significantly improved collective action, except in Shanxi. Among the three groups of agricultural income, the formal institution had the highest impact on the low agricultural income group. A possible explanation is that these farmers have a small agricultural operation scale. Their low dependence on irrigation facilities determines their low enthusiasm for participating in collective action. Under this circumstance, the formal institution can play an incentivizing and constraining role to motivate these farmers to participate in collective action. Of the three groups of provinces, the impact in Ningxia was the greatest. A possible explanation is that the WUA in Ningxia has a relatively high degree of development.

According to our survey, there are two main types of institutional arrangements for irrigation system operation and maintenance: a WUA-dominating arrangement and a village committee-dominating arrangement. In part of the surveyed areas, WUA is primarily responsible for managing and maintaining irrigation systems at the village level. These WUAs normally cooperate with local village committees to perform their responsibilities. On the other hand, in areas without WUAs, the local village cadres normally organize villagers to complete the relevant work. However, in some special areas, like most parts of Inner Mongolia, it is very hard for villagers themselves to directly engage in irrigation management and maintenance work because their irrigation areas and irrigation facilities are relatively large. In these areas, village cadres will collect fees from villagers and purchase services from third parties.

In Ningxia, the reform of the agricultural water supply management system and the water price formation mechanism was carried out in 2005. The model, which involves WUA managing water facilities, has been fully implemented, emphasizing the role of the association in connecting the water management units and the vast number of farmers. Furthermore, the WUA can implement the irrigation management institution together with the village committee, thus motivating farmers to actively participate in irrigation management.

Thirdly, the formal institution had an obvious positive effect on moral obligation, except in Shandong. Among the three groups of agricultural income, the difference between groups was small. This shows that the impact of the formal institution on farmers’ moral obligation does not vary significantly with the difference in agricultural income. Of the three groups of provinces, Ningxia’s formal institution had the greatest impact on farmers’ sense of moral obligation. A possible explanation is that the high development level of the WUA in Ningxia provides a strong impetus for the implementation and improvement of the formal institution in the village. This could help farmers gradually internalize the formal institution into their own sense of morality by following the institutional norms. Accordingly, the farmers’ moral obligation and the occurrence of moral behaviors would be improved in irrigation management.

Fourthly, collective action significantly promoted PIM performance in all groups. Among different agricultural income groups, the low agricultural income group has the highest impact. A possible explanation is that farmers with low agricultural income generally have small areas of irrigable farmland and need less management investment, so they can realize the adequate management of irrigation canals in the field via collective actions, thus achieving a high PIM performance. What is more, the impact in the three provinces was high. This shows that collective action in the three provinces can be effectively transformed into PIM performance. Relatively, collective action had a lower effect on PIM performance in Ningxia than that in Shanxi and Shandong. The possible reason might be that the average age of the irrigation management participants in Ningxia is relatively higher. Since irrigation system maintenance normally requires strenuous physical labor, older participants would have a lower level of PIM performance, although WUAs play a stronger role and effectively generate collective action in Ningxia.

5.4. Limitations and Implications for Future Research

Two main limitations might reduce the significance of this study. Firstly, the omission of variables, like informal institutions, gender, social capital, and social norms, might dampen the explanatory power of our model. However, it is infeasible to incorporate most of the main related influencing variables in the same SEM practically. Secondly, the style of questioning by our enumerators and the deliberate responses from our respondents might decrease the accuracy and reliability of the research results to a certain extent. In the future, a more robust theoretical framework should be generated, and the interaction mechanism between moral obligation and social norms should be incorporated into it.

6. Conclusions

This research analyzed the impact of moral obligation and formal institutions on the performance of participatory irrigation management (PIM) in rural China, based on a database covering 712 sample households in the Yellow River basin. A multi-group structural equation model (SEM) was selected to divide the samples into six groups according to the level of peasant household agricultural income and the provinces in which they are located. The main conclusions are as follows. Firstly, both moral obligation and formal institutions can markedly promote the participation of farmers in collective action, and moral obligation has a greater impact. Furthermore, collective action boosts PIM performance, and the formal institution can significantly improve the farmers’ moral obligation. Besides, the support level given by village organizations to farmers in irrigation, which is the control variable, played a positive role in PIM performance. Secondly, the results of the multi-group analysis show that the effects of the moral obligation and formal institutions on collective action and PIM performance vary across peasant household agricultural income levels and provinces. Among different agricultural income groups, the moral obligation of farmers with a high agricultural income had the greatest effect on their participation in collective action. The formal institution of the low agricultural income group had the greatest impact on collective action, and the collective action of this group can be effectively transformed into PIM performance. The formal institution in Ningxia had the greatest impact on moral obligation and collective actions. Collective actions could lead to high PIM performance in all three provinces.

Given the role of moral obligation and formal institutions in improving collective action and PIM performance, we suggest several policy implications. Firstly, moral obligation is an internal constraint mechanism that effectively promotes farmers’ participation in irrigation management. Therefore, moral obligation, as a strong supplement to formal institutions, can help to promote the effective implementation of the self-governance of irrigation facilities. Specifically, governments could actively develop rural grassroots organizations and motivate farmers to communicate better through these organizations. Given that they are based on promoting information-sharing and cooperation among farmers, these organizations and the village committee could work together to gradually internalize the rules of irrigation management into the morality of farmers. On this basis, this is conducive to the formation of an orderly cooperation mechanism, effectively realizing irrigation management. Secondly, the formal institution, as an external incentive and constraint mechanism, can effectively promote the participation of farmers in irrigation management. However, the significant imperfection of the village’s formal institution has seriously affected the effective governance of CPRs. Therefore, the government can promote the improvement of the formal institution of the village in various ways. To be specific, firstly, the village committee needs to generalize the conventional irrigation management rules in the village and formulate them into a formal institution, so that villagers can have a basis to follow when participating in irrigation management. Secondly, the village committee should clarify the responsibilities and obligations of irrigation management, and specify the specific tasks of management to each family. The right of farmers to use irrigation facilities also needs to be clarified to promote a virtuous circle of the use and management of irrigation facilities. Thirdly, the village committee should strengthen the supervision and punishment institution in irrigation management. The village committee could set up special supervisors to strongly supervise and punish free-riding and the violation of irrigation orders in irrigation management, ensuring the fairness and justice of irrigation management.

The main innovation of this research was to investigate the influence of moral obligation and formal institutions on PIM, expanding the research field of CPR governance. Moreover, this research innovatively analyzed the relationship between moral obligation and formal institutions, which had not been examined in previous studies. However, this study still has some shortcomings. Firstly, our research area is the Yellow River basin, which only focuses on irrigation management in northern China, but does not cover southern China, so it cannot fully represent irrigation management in China as a whole. Secondly, the data in this study are cross-sectional and thus do not reflect the changes in PIM over time. It is expected that these deficiencies can be overcome in future studies. Additionally, this study can be further analyzed in future research from the aspects of the impact of land circulation, the adoption of water-saving irrigation technology, and other factors involved in irrigation management.