Factors Affecting Former Fishers’ Satisfaction with Fishing Ban Policies: Evidence from Middle and Upper Reaches of Yangtze River

Abstract

The Yangtze River fishing ban policy is one of the most important ecological protection measures in middle and upper reaches of Yangtze River. Research on fishers’ satisfaction with the ban will allow policymakers to improve and further optimize it. Based on the theoretical framework of sustainable livelihoods, policy cognition variables are used to explore how livelihood capital and policy cognition differences bring about satisfaction disparities. The research area includes three counties and cities in the Chishui River basin of Guizhou Province, together with Honghu City of Hubei Province, which were among the first areas of the country to implement this policy. The ordered probit model and structural equation model were applied and analyzed based on data that were collected through interviewing the fishers affected by the ban. The results indicate the following: (1) Physical capital, human capital, financial capital, and social capital are significantly and positively correlated with fishers’ satisfaction regarding the Yangtze River fishing ban. In contrast, natural capital does not significantly impact satisfaction. (2) Livelihood capital types have different impacts on the satisfaction of fishers regarding policies for the last aspects. The influence order has the following sequence: financial capital, physical capital, human capital, and social capital. (3) Enhancing fishers’ understanding of the ban could enhance their satisfaction with it. While formulating compensation policies, the government should comprehensively consider the impacts of livelihood capital, formulate special policies to perfect legislation and social security, and use more effective public relations strategies to raise fishers’ awareness of withdrawal policies. Notably, the selected variables and methods in this paper have the potential to significantly enhance the existing literature in the field of ecological management.

1. Introduction

With the development of the Chinese economy in recent years, many intensive human activities have brought both huge economic benefits and great changes to the aquatic ecological environment of the Yangtze River. The living environment of aquatic organisms in the Yangtze River is deteriorating, and the index of biodiversity continues to decline. Moreover, rare endemic species numbers are greatly decreasing [1]. From 2017 to 2024, policies such as The First Document issued by the Central People’s Government of the People’s Republic of China, the Yangtze River Protection Law (2021), and the Ecological Protection Red Line Initiative (2017) stressed the need to implement a no-fishing policy in the Yangtze River basin; accordingly, the no-fishing policy became an important institutional demand. The withdrawal of fishers from fishing to curb the decline in aquatic living resources in the Yangtze River is necessary to protect the regional ecological environment [2]. In this way, the control of resources and ecology in the Yangtze River basin can be carried out in more broad-spectrum terms. The ecological Compensation System, a new paradigm in ecological management, employs economic regulation as a means of operation, with law serving as a guarantee, and focuses on the remediation and restoration of the ecological environment. Ensuring that fishers receive economic benefits through ecological compensation is the best way to balance economic development within the basin with the protection of the ecological environment [3]. Nearly 230,000 fishers from 10 provinces and cities along the Yangtze River lost their original means of livelihood after the ban on fishing. This measure has also posed new challenges for poverty alleviation and the building of a moderately prosperous society from all angles. The Chinese government also proposed the development strategy of the Yangtze River Economic Belt, with “ecological priority and green development” as the core concept, to solve the problem of over-drawing the ecological environment of the Yangtze River. Beginning in 2021, a 10-year ban will be implemented on productive fishing in the mainstream of the Yangtze River and its major tributaries. Recovery of fishing gear, boats, and fishing licenses in the form of cash compensation has been pursued through a compensation mechanism developed by local governments and central governments to ensure the livelihoods of fishers who have retired from fishing. According to the theory of feasible abilities, farmers’ livelihoods, which rest on various feasible functional activities, provide them with the freedom to achieve a better quality of life [4]. After the fishing ban, the livelihoods of fishers will be different, and this will directly affect the satisfaction of fishers with the policy. Moreover, the satisfaction of fishers with the fishing ban will also be affected by the conservation of the Yangtze River and the sustainable development of conservation causes.

Livelihood capital is the key to assessing the sustainable livelihoods of fishers [5]. To date, many scholars have researched the transformation in the livelihoods of fishers after landing [6,7,8,9] but rarely have they analyzed the satisfaction of fishers from the perspective of livelihood capital. In the existing literature on factors affecting farmers’ policy satisfaction, empirical research was conducted using different methods, including a fuzzy set qualitative comparative analysis, customer satisfaction model, TOPSIS method with improved entropy weight, structural equation model, ordered probit model, and logit model, to assess the impact of each observed variable on satisfaction [10,11,12,13,14,15,16]. Therefore, this paper referred to the methods of other papers, and first used the ordered probit model to analyze the observed variables’ influence on satisfaction with the fishing ban and screened out the significant variables to improve the study’s reliability. The analysis’ results were used as a basis for assessing satisfaction related to livelihood capital and policy cognition based on the fishing ban; these factors were analyzed via a structural equation model to give the results a scientific basis, ensuring both the livelihoods of fishers and the implementation of the fishing ban policy.

Previous research has pointed out that livelihood capital, such as human, natural, physical, financial, and social capital, plays an important role in determining satisfaction with public policies. However, the interaction of these forms of capital, or the influence of policy recognition on satisfaction, is often excluded from most analyses. This paper attempts to fill this gap by using sustainable livelihood frameworks and structural equation models to sort out the complex relationship between livelihood capital, policy perception, and satisfaction. Meanwhile, feasible policy recommendations are made to enhance the socioeconomic resilience of fishers in order to support ecological and social policy objectives.

2. Data and Methodology

2.1. Theoretical Analysis and Research Hypotheses

Fishing is one form of modern agriculture, but its relatively few participants could easily lead to fishers being marginalized in society and becoming a vulnerable group [17]. After the increase in the protection of the Yangtze River basin by the government, when fishers withdraw from the natural and ecological environment, they will permanently lose their traditional production relationship and means of production. This huge change in their livelihood capital puts fishers’ sustainable livelihoods at risk. The concept of sustainable livelihood was devised by the British Department for International Development (DFID) through the sustainable livelihood analytic framework. It categorizes livelihood capital into five types, human capital, natural capital, physical capital, financial capital, and social capital, describing how fishers use property, rights, and strategies to cope with the risk environment and sustain and develop their businesses [18]. For example, Atmis et al. (2007) observed that when motivating farmers to implement measures for protecting farmland, whether or not they were satisfied with the policy would affect their ecological behavior [19]. Defrancesco et al. (2008) went further and pointed out that, besides the influence of individual factors, such as family income, personal attitudes, and beliefs, farmers’ participation in agricultural and environmental projects may influence their satisfaction with the policies [20]. Many scholars have researched the factors affecting policy satisfaction from the perspective of livelihood capital. Huang Zhigang et al. (2018), Xie Jin et al. (2017), and Wang Changhai et al. (2017) agreed that based on the sustainable livelihood framework, various livelihood capitals have different impacts on farmers’ life satisfaction [13,21,22]. Studies by Han Feng et al. (2017) and Liu Di et al. (2018), while observing the satisfaction of herders/farmers with ecological policies, identified a direct relationship between both herders’ understanding of grassland ecological protection policies and farmers’ understanding of policy objectives and the level of satisfaction with the policies [14,23]. For example, Hounsome et al. (2006) found that, in terms of individual farmers, differences in understanding and expectations regarding the policies led to different levels of satisfaction regarding the compensation policies for cultivated land protection [24]. In another study, Liu et al. (2018) found that differences in farmers’ livelihood endowments influenced farmers’ enthusiasm for participating in ecological compensation policies [25].

Based on previous research experience and the actual lives of retired fishers, we apply the theory of sustainable livelihoods to analyze the impacts of the five major livelihood capital variables and policy cognition on fishers’ satisfaction with the fishing ban.

Physical capital denotes the infrastructure and means of production required for sustenance. During the ban-related process of resettlement, all fishers will live on land and use their compensation money to cover housing, transportation, household appliance, and production equipment costs. Their material capital will thus increase, their living standards will improve, and their satisfaction levels with the ban will increase. Therefore, the following hypothesis is constructed:

Hypothesis 1 (H1):

The material capital owned by fishers has a positive significant influence on their satisfaction with the fishing ban policy [6,8].

Human capital represents the skills that workers use to complete productive labor, including knowledge, labor ability, and health level. The more human capital a fishing family has, the stronger its intrinsic productivity, so the use of the other four kinds of capital ability will also increase. That means that they will have more opportunities to access high-income-generating jobs, making them more satisfied with the fishing ban policy. Thus, Hypothesis 2 is constructed:

Hypothesis 2 (H2):

The human capital possessed by the fishers has a positive significant influence on their satisfaction with the fishing ban policy [5,6].

Financial capital represents the future cash flows that are available to fishers. The most immediate sense of satisfaction with any policy is the increase in income and the ability to withstand the external risks that it provides. Compensation becomes the income of fishers and forms their financial capital. The more money a fisher has, the more financial capital they have and the more satisfied with the fishing ban they become. Thus, Hypothesis 3 is constructed as follows:

Hypothesis 3 (H3):

The financial capital held by fishers has a positive significant influence on their satisfaction with the fishing ban [2,7].

Social capital represents the social resources available to fishers that they can use to meet a certain goal in life. After withdrawing from fishing, social relationships between fishers and village cadres, neighbors, and communities as a whole can be constantly enhanced, individual ability can be promoted, the sense of integration into the new living environment can be heightened, and, consequently, satisfaction with the fishing ban policy can be improved. In this context, Hypothesis 4 is drawn up as follows:

Hypothesis 4 (H4):

The social capital possessed by fishers has a positive significant influence on their satisfaction with the fishing ban policy [8,12].

Natural capital is defined as the natural resources that can be used, along with the environmental services that are provided by them. To produce natural capital, fishers use physical capital and human capital. While owning the same amount of human capital and physical capital, with more services being provided by natural capital, fishers’ outputs will be higher and, thus, their incomes will rise; accordingly, fishers’ satisfaction with the policy will be higher. Therefore, Hypothesis 5 is proposed as follows:

Hypothesis 5 (H5):

The natural capital owned by fishers has a positive significant influence on their satisfaction with the fishing ban policy [6,9].

Policy awareness explains fishers’ understanding of a policy. The higher the level of awareness of the policy among fishers, the more likely they will accept the fishing ban policy as being in the long-term interests of society and the higher their satisfaction with the policy. Therefore, Hypothesis 6 is formulated as follows:

Hypothesis 6 (H6):

The degree of awareness among fishers regarding the ban policy has a positive significant influence on their satisfaction with the ban policy [10,14].

2.2. Research Design

2.2.1. Data Source

The data in this paper were derived from a questionnaire survey of fishers conducted in January 2019 in 3 counties and cities of the Chishui River basin of Guizhou Province and in July 2019 in Honghu City, Hubei Province. In 2017, the No. 1 Central Document proposed to strengthen the fishing and time limits in all areas of the river, lake, and sea and recommended a total ban on fishing in the aquatic life protection area of the Yangtze River basin. An important section of the Yangtze River—the Chishui River—implemented a full fishing ban in 2016. The National Nature Reserve in Honghu City implemented a full ban at the beginning of 2018, generating representative data. To ensure the continuity of the data, both places used the same questionnaire.

The survey was divided into two parts. Information on fishers’ withdrawal from fishing and changes in production and operation was first obtained from the county departments, the towns under investigation, and other related units. This paper discusses the basic contexts of local implementation, with departments being in charge of designing and formulating policies for withdrawal. Secondly, through the adoption of methods such as questionnaire surveys and household interviews, the researchers used one-on-one question-and-answer methods, thereby obtaining subsistence capital assessment data for fishers. All in all, 200 questionnaires were distributed, from which 179 valid questionnaires were retrieved, with the effective sample rate being 89.5%. The survey objects were all professionals or part-time fishers who were withdrawn from fishing in the Yangtze River basin, excluding aquaculture fishers. Referring to Wang Changhai (2017) [13] and Yuan Liang (2017) [26], the questionnaire was designed to fit this research topic. The detailed design variables are presented in

Table 1. Variable index.

Variable	Variable Meaning and Assignment	Mean	Standard Deviation
1. Physical capital (P)
Per capita residential area (P1)	The ratio of house size (square meters) to family size: P1 ≤ 10 (1); 10 < P1 ≤ 20 (2); 20 < P1 ≤ 30 (3); 30 < P1 ≤ 40 (4); P1 ≥ 40 (5)	3.120	1.088
Types of durable household goods (P2)	Types of durable goods owned: P2 = 0 (1); P2 = 1–2 (2); P2 = 3–4 (3); P2 = 5–6 (4); P2 ≥ 7 (5)	3.260	0.925
Transportation price (P3) (CNY)	Total purchase price of vehicles (cars, motorcycles, electric vehicles, etc.): P3 ≤ 10,000 (1); 10,000 < P3 ≤ 30,000 (2); 30,000 < P3 ≤ 50,000 (3); 50,000 < P3 ≤ 80,000 (4); P3 ≥ 80,000 (5)	2.930	1.047
2. Human capital (H)
Family education level (H1)	Uneducated (1); primary school (2); junior high school (3); high school (4); college degree or above (5)	2.740	1.035
Labor force (H2)	H2 = 0 (1); H2 = 1–2 (2); H2 = 3–4 (3); H2 = 5–6 (4); H2 ≥ 7 (5)	3.060	1.150
Family health (H3)	Long-term illness (1); frequent illness (2); occasional illness (3); rarely get sick (4); not sick (5)	3.440	1.176
3. Financial capital (F)
Annual household income (F1) (CNY)	F1 ≤ 10,000 (1); 10,000 < F1 ≤ 30,000 (2); 30,000 < F1 ≤ 50,000 (3); 50,000 < F1 ≤ 80,000 (4); F1 ≥ 80,000 (5)	2.990	1.119
Loan situation (F2)	Is it easy to obtain a loan: Very difficult (1); difficult (2); average (3); easy (4); very easy (5)	3.040	1.032
Insurance (F3)	Types of health insurance and pension insurance: F3 = 0 (1); F3 = 1 type (2.5); F3 = 2 types (5)	2.660	1.341
4. Social capital (S)
Relationships with relatives and friends (S1)	Contact time with relatives and friends: Barely (1); low (2); average (3); much (4); high (5)	3.180	0.951
Relationships with village cadres (S2)	Contact times with village cadres: Barely (1); low (2); average (3); much (4); high (15)	3.070	0.936
Number of times of participation in village (community) activities (S3)	S3 = 0 (1); S3 = 1–2 (2); S3 = 3–4 (3); S3 = 5–6 (4); S3 ≥ 7 (5)	2.710	1.046
5. Natural capital (N)
Cultivated area (N1) (mu) (1 ha = 15 mu)	N1 = 0 (1); 0 < N1 ≤ 1 (2); 1 < N1 ≤ 3 (3); 3 < N1 ≤ 5 (4); N1 ≥ 5 (5)	2.110	1.080
Aquaculture area (N2) (mu) (1 ha = 15 mu)	N2 = 0 (1); 0 < N2 ≤ 1 (2); 1 < N2 ≤ 3 (3); 3 < N2 ≤ 5 (4); N2 ≥ 5 (5)	1.210	0.716
6. Policy cognition (EC)
Knowledge of the policy (EC1)	Very little understanding (1); do not know much about it (2); general understanding (3); good understanding (4); understand it very well (5)	3.110	1.059
The necessity of policy implementation (EC2)	Very unnecessary (1); unnecessary (2); neutral (3); necessary (4); very necessary (5)	3.400	0.997

.

The analysis and comparison of the mean based on the observed variables of all hidden variables showed that the average level of policy cognition is the most common. The average of the relevant variables is over 3.0. The lowest is natural capital, which had an average of less than 2.2.

2.2.2. Descriptive Statistics of the Measurement Results of Satisfaction

According to

Table 2. The distribution of fishers’ satisfaction with the Yangtze River fishing ban policy.

Satisfaction	Very Dissatisfied (1)	Dissatisfied (2)	General (3)	Satisfied (4)	Very Satisfied (5)	Mean	Total
Total sample	8 (4.47%)	40 (22.35%)	64 (35.75%)	51 (28.49%)	16 (8.94%)	3.15	179

, the fishers’ satisfaction with the ban policy in the Yangtze River shows a clear distribution of data; the descriptive statistical result is shown in this table, where ‘general’ accounts for 35.75% of all samples surveyed and has the largest proportion, ‘dissatisfied’ accounts for 22.35%, ‘satisfied’ accounts for 28.49%, and the rest is ‘very dissatisfied’ and ‘very satisfied’, the proportions of which account for 4.47% and 8.94%, respectively. Overall, based on the average satisfaction (3.15), the satisfaction of the sample fishers is between ‘neutral’ and ‘satisfied’ but closer to ‘neutral’.

2.2.3. Test of Questionnaire Data Reliability and Validity

The reliability and validity tests of the data from the questionnaires are presented in this study. After calculation, the Cronbach’s alpha of the questionnaire data was 0.886, which is greater than 0.8. Therefore, the intrinsic reliability of the evaluation system was relatively good on the whole. The validity analysis of the questionnaire results was as follows: the KMO value was 0.925, and the common degree of each measurement variable was over 0.5, meeting the desired standard. So, the measurement variables of the questionnaire designed in this paper have good validity, indicating that the designed measurement variables of the questionnaire are ideal.

3. Empirical Models and Results of Fishers’ Satisfaction with the Fishing Ban Policy

The quantitative influence factor of fishers’ satisfaction with the fishing ban policy and how livelihood capital influences satisfaction with the fishing withdrawal policy were studied based on the selective variable using the econometric model. In this study, ordered probit models were applied to filter out observable variables having significant effects on fishers’ satisfaction. Then, using livelihood capital and policy cognition as exogenous hidden variables and policy satisfaction as an endogenous hidden variable, the structural equation model was used to analyze the influence mechanisms of observed variables and exogenous hidden variables on endogenous hidden variables. In this way, the structural equation model was more targeted and significant when analyzing the influencing mechanisms of observed variables and exogenous hidden variables on endogenous hidden variables.

3.1. Empirical Models of Ordered Probit and the Structural Equation Model (SEM)

The basic settings of the ordered probit model were as follows:

$${{Satisfaction}^{*}}_i={\alpha }_0+{\alpha }_1{Livelihood\_Assets}_i+{\alpha }_2{EC}_i+{\epsilon }_i$$

The dependent variable was denoted as Satisfaction^*_i, which represents the satisfaction level of fishers. The following question was designed to construct a fisher satisfaction index regarding the Yangtze River fishing ban: ‘In general, are you satisfied with the Yangtze River fishing policy?: 1 = very dissatisfied; 2 = not satisfied; 3 = neutral; 4 = satisfied; 5 = very satisfied.’ Among the explanatory variables, ${Livelihood\_Assets}_i$ represents all kinds of quantitative standard livelihood capital owned by the family, including material capital (P_i), human capital (H_i), financial capital (F_i), social capital (S_i), and natural capital (N_i). ${EC}_i$ is the perception of the policy, $\alpha$ is the regression coefficient, and ${\epsilon }_i$ is a random error term.

The structural equation model focuses on the causality relationship between hidden variables, which can effectively overcome the collinearity between variables. The PLS structural equation modeling tool is a kind of soft SEM modeling method that does not assume the distribution of data [27]. For small samples, when the prediction accuracy is of critical importance, the partial least square structural equation model (PLS-SEM) was used as a substitute for the covariance-based structural equation model (CB-SEM) [28]. The estimation method is quite specific and described in the following equation:

$$\begin{array}{c}\mathrm{Structural} \mathrm{equation}: Y={\gamma }_{1}P+{\gamma }_{2}H+{\gamma }_{3}F+{\gamma }_{4}S+{\gamma }_{5}N+{\gamma }_{6}EC+\mu \\ \mathrm{Measurement} \mathrm{equation}: Y={\lambda }_{y}Y+\epsilon \\ P_i={\lambda }_{pi}P+{\delta }_{pi} (i=1,2,3) H_i={\lambda }_{hi}H+{\delta }_{hi} (i=1,2,3)\\ F_i={\lambda }_{fi}F+{\delta }_{fi} (i=1,2,3) S_i={\lambda }_{si}S+{\delta }_{si} (i=1,2,3)\\ N_i={\lambda }_{ni}N+{\delta }_{ni} (i=1,2) {EC}_i={\lambda }_{eci}EC+{\delta }_{eci} (i=1,2)\end{array}$$

In contrast, in the structural equation, Y is taken as an endogenous hidden variable employed for identifying the satisfaction of fishers. Moreover, exogenous hidden variables such as P, H, F, S, N, and EC represent material capital, human capital, financial capital, social capital, natural capital, and policy cognition, respectively. ${\gamma }_p, {\gamma }_h, {\gamma }_f, {\gamma }_s, {\gamma }_n, \mathrm{a}\mathrm{n}\mathrm{d} {\gamma }_{ec}$ represent the path coefficients between the hidden variables; $\mu$ is the unexplained part of the endogenous hidden variable Y. The measurement equation reflects the relationship between the hidden and observed variables. In the measurement equation, ${\lambda }_p, {\lambda }_h, {\lambda }_f, {\lambda }_s, {\lambda }_n, \mathrm{a}\mathrm{n}\mathrm{d} {\lambda }_{ec}$ represent the factor load between the exogenous hidden variable and the observed variable, ${\lambda }_y$ represents the factor load between the endogenous hidden variable and the observed variable, $\delta$ represents the measurement error of each observed variable of the exogenous hidden variable, and $\epsilon$ represents that of the endogenous hidden variable. The research structure is represented in Figure 1 and the framework used in the analysis is illustrated in Figure 2.

However, to test the hypothesis, it was necessary to analyze the measurement model and test the reliability and validity of each constructed variable. This paper applied the PLS structural equation tool SmartPLS 3.2 to conduct a bootstrapping estimation for the significance calculation of each factor loading and path coefficient.

3.2. Results of Fishers’ Satisfaction with the Fishing Ban Policy

The empirical results of the ordered probit model are shown in

Table 3. Ordered probit model regression results.

Variable	Coefficient	Standard Error	Z Value
Per capita residential area, P1	0.515 ***	0.158	3.260
Durable household goods, Category P2	0.465 **	0.188	2.470
Transportation price, P3	0.338 **	0.158	2.140
Family education level, H1	0.403 **	0.177	2.280
Labor force number, H2	0.121	0.159	0.760
Family health status, H3	0.447 ***	0.135	3.310
Annual household income, F1	0.501 ***	0.151	3.320
Credit situation, F2	0.712 ***	0.176	4.060
Insurance, F3	0.350 **	0.147	2.390
Relationships with relatives and friends, S1	0.484 **	0.199	2.430
Relationships with village cadres, S2	−0.151	0.192	−0.790
Number of activities organized in the village, S3	0.101	0.167	0.610
Cultivated land area, N1	−0.035	0.123	−0.290
Culture area, N2	−0.121	0.172	−0.710
Policy understanding, EC1	0.258	0.165	1.560
Necessary degree of policy implementation, EC2	0.833 ***	0.206	4.040
LR chi2 (16)	379.970
Prob > chi2	0.000
Pseudo R2	0.750

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

.

The results in Table 3 reveal that the regression of the model was significant as a whole, and the goodness of fit was 0.750, which has a good explanation effect. In terms of physical capital, the two variables of residential area per capita (P1) and durable household goods (P2) significantly positively affect satisfaction, underlining the role of improved housing and living standards. Subsidies for housing and durable goods can further raise satisfaction. Improved transportation prices (P3) also affect satisfaction positively, showing the need for transportation support. From a human capital perspective, educated (H1) families were more pleased with the policy, again helping to indicate that outreach education will increase policy acceptance. Healthy (H3) families exhibited more life satisfaction, further indicating that access to healthcare during times of transition remains paramount. From the point of view of financial capital, stability in annual family income (F1) is important for satisfaction. Compensation schemes should provide adequate economic support to address economic challenges. The variable of credit access (F2) has the highest coefficient, implying that credit availability is critical to support the transition in alternative livelihoods of fishers. Health and pension insurance (F3) participation contributes positively to satisfaction, signaling a sound social security system. Strong social networks (S1) enhance satisfaction; therefore, community-building initiatives that increase social integration may be more acceptable to policy. In policy recognition, there was an overbearing weight from the perceived need to ban fishing, thus calling for transparent communication and public awareness. Natural capital (N1, N2) had no significant effect; it expressed its limited significance for professional fishers. Therefore, a structural equation model was used to further test the categorical variables of observable variables that influenced the significance.

Table 4. Measurement variable load matrix.

	F	H	P	S	EC	y
F1	0.84	0.437	0.536	0.397	0.49	0.664
F2	0.795	0.516	0.57	0.444	0.528	0.702
F3	0.701	0.445	0.389	0.45	0.38	0.538
H1	0.607	0.871	0.623	0.523	0.478	0.697
H3	0.373	0.801	0.488	0.486	0.358	0.571
P1	0.424	0.543	0.724	0.462	0.453	0.606
P2	0.521	0.434	0.707	0.4	0.564	0.594
P3	0.492	0.511	0.788	0.56	0.44	0.645
S1	0.547	0.603	0.643	1	0.496	0.694
EC2	0.603	0.505	0.653	0.496	1	0.734
Y	0.819	0.763	0.83	0.694	0.734	1

Note: P—material capital; H—human capital; F—financial capital; S—social capital; EC—policy cognition; Y—withdrawal satisfaction.

and

Table 5. Descriptive statistics of variables and correlation coefficients between variables.

	Mean	S.D.	CR	AVE
F	0.342	0.042	0.823	0.610
H	0.219	0.053	0.823	0.700
P	0.274	0.053	0.784	0.549
S	0.105	0.039	1.000	1.000
EC	0.184	0.042	1.000	1.000

Note: P—material capital; H—human capital; F—financial capital; S—social capital; EC—policy cognition.

show the results of the variables’ validity and reliability analyses. The factor loadings of each measurement index in this study were more than 0.7 and, thus, significant. Gefen et al. believe that when the factor loadings of every measurement index for the corresponding construction variable are above 0.7 and significant, the scale has great convergent validity [29]. In general, these results showed the convergent validity of the indexes of measurement. Meanwhile, the average variance extracted (AVE) from all measured variables was greater than 0.5. Thus, we could ensure that each measurement index had good convergent validity. In Table 5, we can see that the composite reliability (CR) values for all variables were larger than 0.7, indicating the constructed good reliability of each variable.

The factors that had a significant impact on fishers’ satisfaction when withdrawing from fishing are shown in Table 3. On this basis, a structural equation model was further used to test whether the categorical variables of the observable variables with significant influence, namely hidden variables such as financial capital, material capital, human capital, social capital, natural capital, and policy cognition, had overall significance for fishers’ satisfaction with the fishing ban. In this study, SmartPLS 3.2 was used to verify the structural model of the influencing factors of fishers’ satisfaction with the fishing ban. The results of the regression are shown in

Table 6. The calculation results of the resulting model.

Path	Influence Direction	Path Coefficient	Distinctiveness	Test Result
Financial capital → Withdrawal satisfaction	+	0.342 ***	0.000	Support H1
Human capital → Withdrawal satisfaction	+	0.216 ***	0.000	Support H2
Material capital → Withdrawal satisfaction	+	0.273 ***	0.000	Support H3
Social capital → Withdrawal satisfaction	+	0.108 ***	0.005	Support H4
Policy cognition → Withdrawal satisfaction	+	0.187 ***	0.000	Support H6

Note: *** indicates significance at 1%. + indicates that the variable has a positive effect on satisfaction.

. For the PLS path model, the measurement coefficient R² was 0.884, indicating that the six factors could explain fishers’ satisfaction with fishing very well.

As shown in Table 6, among the five variables that passed the significance test, the path coefficient of fishers’ satisfaction upon returning from fishing was high, thus indicating that the explanatory power of these five variables for fishers’ satisfaction upon returning from fishing was good. The ranking order of livelihood capital variables for explaining the satisfaction of fishers towards fishing was financial capital, material capital, human capital, and social capital. Fishers’ satisfaction was significantly influenced by policy cognition, but it did not affect natural capital. Based on the above results, all the hypotheses, except for Hypothesis 5, were supported.

4. Discussion of Empirical Results

(1)

In this study, we assessed the influence of financial capital on fishers’ satisfaction with the fishing ban. In both Table 3 and Table 6, we can see that the path coefficient for the way that the hidden variables of financial capital influence fishers’ satisfaction with the fishing ban is maximal at 0.342. Among them, the household annual income, loan access, and insurance access had significant positive impacts on fishers’ satisfaction with the fishing ban at p < 0.01, p < 0.01, and p < 0.05, respectively. One-off cash compensation is among the government’s publicized ecological compensation policies, including compensation for the loss of fishing rights, compensation for the collection of fishing gear, and transitional living allowances. Compensation directly increases the income levels of fishers in the short term. The measures also include preferential micro-credit policies that aim to solve the financial constraints placed on fishers forced to change industries. To reduce some risks resulting from the fishing ban, some local governments have purchased insurance for the fishers. One thing is obvious: increasing the financial capital of fishers will enhance their satisfaction with the fishing ban policy [2,26].

(2)

The effect of the hidden variable of material capital on the satisfaction of fishers with the fishing ban is significant at a 1% level, with its path coefficient being 0.273. Simultaneously, the per capita residential area, durable household consumer goods, and transportation price have significant positive effects on fishers’ satisfaction with the policy at significance levels of p < 0.01, p < 0.05, and p < 0.05, respectively. The policy influencing fishers included professional fishers and part-time fishers. Some of the fishers have lived on their boats for a long time without having any housing on shore. Under the withdrawal policy, these fishers will return to live on the shore again. After landing, some fishers will surely purchase household necessities so that they can live in a town or county. Therefore, the housing area and durable household goods will have a major positive impact on satisfaction with the fishing ban policy. By directly improving the quality of life of a fisher’s family, material capital will make fishers more willing to actively accept the change in the fishing policy [1,8].

(3)

Considering human capital as a hidden variable, at the 1% level, it has a considerable effect on fishers’ satisfaction, with the path coefficient being 0.216. Two of the observable variables categorized as human capital—the education level of their family (at a level of 0.05) and the status of their family’s health (at a level of 0.01)—have significant positive correlations with fishers’ satisfaction. The higher the level of education, the more fishers can realize the importance of protecting the ecological environment and are likely to approve the establishment of the exit policy. The poorer the health status of relatives of fishers, the greater the living pressure, reducing satisfaction with the policy [6,23].

(4)

The hidden variable of social capital considers the influence that relationships with relatives and friends have on fishers’ satisfaction with the ban. The path coefficient is 0.108, with only getting along with relatives and friends having a significant positive impact on fishers’ satisfaction with the fishing ban (p < 0.05), indicating that fishers who get along with relatives and friends are more likely to accept the change brought on by the fishing ban [12,25].

(5)

Natural capital slightly affects the satisfaction of fishers. The possible reasons for this are as follows: most fishers are professional fishers who possess little arable land. According to the national policy of ‘keeping the land contract relationship stable and unchanged for a long time [30]’, fishers’ families cannot farm arable land if they withdraw from fishing. As ‘the state promotes the green development of the aquaculture industry’, fishers who withdraw from fishing find it difficult to transfer to natural water aquaculture; moreover, only a few fishers can provide aquaculture services in small fishponds. Thus, cultivated land area and aquaculture area do not have any statistically significant effects on the satisfaction of fishers [18].

(6)

Policy cognition’s relationship with the satisfaction of fishers who withdraw from fishing is significant at a 1% level, and the path coefficient is 0.187, as shown in Table 3. Besides policy cognition, the necessity of this policy has a significant influence on fishers’ satisfaction with the fishing ban, and the test level is p < 0.01. This can be explained by the fact that, before implementing the policy, the government first increased the publicity and explained the necessity of its implementation to fishers [10,24].

5. Conclusions and Policy Recommendations

Using data collected from a questionnaire survey of fishers, ordered probit and structural equation models were used to conduct empirical analyses of the effects of livelihood capital and the cognition of the fishing ban policy on fishers’ satisfaction with that policy. Our findings imply that material capital, human capital, financial capital, and social capital have significant effects on the satisfaction of fishers with the Yangtze fishing ban policy; however, natural capital did not significantly affect the satisfaction of fishers with the fishing ban policy. The ranking of significant factors was as follows: financial capital, material capital, human capital, and social capital. At the same time, this order was affected by fishers’ cognition of the ban, which influenced their satisfaction with the policy. All the observable variables reflecting the hidden variables—the per capita residential area, types of durable household goods, price of transportation, family education level, family health status, family annual income, loan situation, insurance, relationship with relatives and friends, and necessity of policy implementation—all have significant positive effects on fishers’ satisfaction with the ban.

Based on the above research conclusions, the following policy implications were identified:

(1)

Amending legislation to protect the rights and interests of those fishers withdrawing from fishing: Fishers who withdraw from fishing are contributors to the protection of the Yangtze River. Further, in the ensuing Fishery Law, along with the Yangtze River Protection Law, retired fishers should receive legal assurance for their livelihoods. Protecting the aquatic ecology of the Yangtze River basin will then increase the satisfaction of the retired fishers.

(2)

The government should formulate policies that focus on the differences between fishers’ livelihood capacity and livelihood capital. For older fishers with lower livelihood capacity, the government should provide jobs that do not need high vocational skills to solve subsequent livelihood problems. For youth, vocational skills training should be provided to help them transfer into new jobs. For the self-employed fishers establishing small and micro-businesses, preferential policies such as discount loans and tax relief should be given. At the same time, the government should visit the fishers who change production methods and business models and organize community activities to allow them to integrate into their new living environment.

(3)

The government should provide insurance and improve social security. Most of the retired fishers are older and have a low level of education, and it is difficult for them to move into other industries after withdrawing from fishing. Rather than job training and support, the most important issues for these fishers are old-age security, medical care, and transitional-period life security policies. We recommend using compensation funds for the return of fishing and conversion of production and continuing to provide social security and medical insurance for older fishers; for local governments with difficulties supporting special groups, the government could help them to improve the procedures for handling subsistence allowances.

(4)

Increase publicity and social participation: Among the Yangtze River basin’s fishers, there is a lack of means to make a living after withdrawing from fishing; thus, many are likely to return to fishing. Moreover, poaching will likely increase in line with the recovery of fish stocks in the Yangtze River basin. The government should use all kinds of media to publicize the need and rationale for the fishing ban and change the optimism bias of some fishers. Meanwhile, social participation should be increased as much as possible. A fishery supervision and management process involving all aspects of society should be established.

The Yangtze River Basin Withdrawal Policy was officially proposed in 2018 and implemented in 2019. At the time of the survey, only a few areas had completed the fishing withdrawal policy, and limited research results and data were available. Some indications of livelihood capital did not reflect the change in fishers’ capital before or after fishing withdrawal. Therefore, the research primarily targets retired fishers in the basin of the Chishui River in Guizhou Province and Honghu City in Hubei Province. Targeting these provinces, hence, will not intend the coverage of more geographic territories but other central provinces within the Yangtze Basin. Second, we could expand representative diversity in the research scope around the Yangtze River basin to catch on to various influences of policies across different socioeconomic contexts.

At the same time, this paper uses an ordered probit model and SEM. However, it does not adopt qualitative research methods that combine in-depth interviews or case studies to explore the micro-policy implementation process. Therefore, in the following research, we can combine the quantitative and qualitative methods and obtain more detailed data through interviews or focus group discussions to help explain the reasons for statistical results.