Factors Influencing the Roles of Environmental Non-Governmental Organizations (ENGOs) on Environmental Bargaining in Yunnan, China

Abstract

Environmental pollution has become a serious problem in China due to the development of industrialization and urbanization since the reform policies and opening of the economy. Nowadays, ENGOs function as a third party for environmental protection through various awareness and bargaining activities. This study aims to analyze the role of ENGOs in environmental bargaining (EB) and the influencing factors by combining the EB theories. A structural equation model of ENGOs participation in EB was established to compare the role of ENGOs in bargaining the “Kunming PX incident” and the “Nujiang dam incident” in Yunnan, China. The findings show that complex powers and interests play a vital role during EB. The relationship network, media, and human resources are among the most significant factors influencing the role of ENGO out of all the other factors such as funding, openness, knowledge, scale and experience. The strength of ENGO relationship network is crucial for solving environmental problems. This study also suggests that in order for ENGOs to effectively engage in EB, they should be placed in the proper context in the negotiating process. It is necessary to set up efficient public involvement platforms and processes for effective EB.

1. Introduction

The achievement of sustainable development goals (SDGs) has become a global need [1]. Since the middle of the 20th century, environmental pollution issues have become increasingly serious as a result of the advancement of technology and human activities. This has resulted in the rise of environmental non-governmental organizations (environmental NGOs, here referred as ENGOs). ENGOs are non-profit-making entities formed by citizens of civil societies for various societal and humanitarian purposes, functioning autonomously from governments [2]. ENGOs have a main purpose of environmental protection and provide public environmental welfare societal services, which were originally from the United States, Europe and other developed countries [3]. These countries have a strong awareness of civil rights and the public has a high awareness of environmental protection, hence there are more famous ENGOs in the international community. Their scale, influence and financing ability are much bigger than the ENGOs in developing countries. These ENGOs have absorbed many social members to engage in various environmental protection undertakings of their organizations, and gradually formed a strict organizational system with perfect financing channels. ENGOs began to become the third force standing side by side with the government and the market and achieved good results because of their public welfare and professionalism.

ENGOs began to play an increasingly functional role in China’s national governance since the Reform and Opening-Up in 1978, where their number and scale have expanded since then. The Chinese government gradually shifted its administrative logic towards ENGOs from corporatism to collaborative governance [4]. ENGOs that have emerged since the early 1990s are among the most frequently studied NGOs in China [5,6,7,8]. As China’s economy has earned amazing growth in the past three decades, environmental pollutions and ecological damages have also posed daunting threats to the future of the country [9]. In China, ENGOs are also called as “social groups”, which are legitimate organizations with administrative permission of government. ENGOs are different from administrative institutions of government and various economic organizations. These ENGOs are social service organizations such as public welfare charities and grassroots service mass organizations, with the common characteristics of being non-profit, autonomous, non-governmental, voluntary, and organizational.

In the current diversified global environmental governance system, ENGOs do not have the restriction to pursue their own interests due to their worldwide activities and their non-profit and public welfare characteristics. Its influence in the field of global environmental governance and social development has been increasing throughout the years. This is a very distinctive feature of global environmental governance [10]. In recent years, some scholars have focused on ENGO’s involvement in environmental bargaining (EB), i.e., development of a local government-NGO collaborative governance approach to enhance public participation and respond to state decentralization and rising environmental issues in urban areas [11]. Pacheco-Vega and Murdie [12] have developed earlier theoretical work and concluded that the influence of ENGO relies on the ability of local citizens to participate in the advocacy and the vulnerability of the state to external pressure. Villo, et al. [13] proposed a theory that introduces two concepts: shadow-boxing and homeostasis to study the social movements and the interactions among social movements, ENGOs, and the authoritarian state [14]. Yekini, et al. [15] extended research on environmentalism and the NGOs’ roles in promoting environmental accountability. The Kurdish environmental activists, organised in ENGOs, are the protectors of Kurdistan’s natural environment. Hassaniyan, et al. [16] focused on the emergence and objectives of environmental activism in Rojhelat of Eastern Kurdistan, arguing that the growth of environmental activism as ENGOs is a nascent trend in Rojhelat, and requires more in-depth studies of its various aspects.

For the past 30 years, China’s environmental policies have been evolving and deepening in status from national basic policy to sustainable development strategy. The focus has changed from pollution control to a combination of pollution control and ecological protection [17]. Compared with the rich practical activities of ENGOs in China, its theoretical basic research is slightly limited. In recent years, some research on the concept, characteristics, classification, status, and function of ENGOs have been conducted in China [18], developing the problems and countermeasures, modes, and mechanisms of ENGOs’ participation in the governance of the ecosystem in the long run. ENGOs are mainly distributed in the coastal, northwest, and southwest regions of China. In particular, the PX projects are mainly concentrated in developed eastern coastal areas. Thousands of people demanded the government to take effective measures to control and control environmental pollution.

The rise of the concept of sustainable development has also influenced the expansion and collaboration of ENGOs in different industries, which in turn has influenced the participation of ENGOs in corporate social responsibility activities [19]. ENGOs have made unrelenting efforts to disseminate the values and vision of the concept of sustainable development in the Earth Charter to many people around the world, which can enable mankind to develop a just, sustainable and peaceful society in the 21st century [20]. ENGOs themselves attach great importance in collecting environmental information and communicating to the society in order to enhance public awareness of environmental protection. However, most ENGOs rarely put forward specific solutions or suggestions when society and the public face environmental problems, especially in the communication between the government and enterprises. Environmental issues are highly complex, highly specialized, long-lasting and systematic work. At present, ENGOs in China are characterized by significant voluntary nature. Generally, there are no or only a small number of full-time members. Most of the members exist in the form of part-time jobs, internships, and volunteers. In addition, there is a lack of experts, standardized training, and organizational communication.

Mega infrastructures, such as waste incineration plants, P-Xylene (PX) projects, and nuclear power plants, continue to develop, but inevitably bring certain social costs [21]. These projects, also known as Not-in-My-Back-Yard (NIMBY) facilities, have complex social implications [22]. Conflicts caused by NIMBY facilities pose significant threats to urban harmony and sustainability [4,23]. The number of environmental administrative reconsideration initiated by the public is increasing in recent years due to the rise of public participation in China. The enthusiasm of public involvement and supervision of the government’s environmental decision-making is rising, which is helpful for protection of the environment at all levels to get rid of the idea of administrative standard. ENGOs were involved in the Kunming PX incident and the Nujiang Dam incident. For the Kunming PX incident, although there were many opposing opinions, the project was quietly started. In the Nujiang Dam incident, the voices of ENGOs such as Green Home and the “Mass Valley” effectively affected the development of the matter. Green Home can mobilize considerable public opinion and social forces in a very short time to exert pressure and attract the public’s attention. Why do similar historical events have different results?

In the field of environmental protection, there are many theories related to the ENGOs’ research, such as market failure theory [24], third party government theory [25], and Western environmental movement theory [26]. However, there are only a few theoretical studies on ENGOs’ participation in environmental bargaining (EB) at present. EB includes three elements: bargaining power, bargaining strategy and bargaining outcome. It emphasizes the role of civil society, especially ENGOs. It believes that not only the management effectiveness of government should be utilized, but also the role of public participation should be played. Joint governance of environmental issues should be implemented through cooperation, consultation, partnership and the establishment of common goals. However, from the EB theory perspective, through the comparison of typical cases to explore the role of ENGOs in the interactive relationship with the public, the government in place and the enterprise, as well as the influencing factors of its role and other related research are very few. Therefore, it is important to understand the factors influencing the current role of ENGOs in order to optimize their roles.

The definition of EB varies from country to country. The German environmental protection agency defined EB as “a binding bargaining or a non-binding bargaining in the economic field designed to protect the environment, in order to achieve environmental goals, to reduce environmental burden activities or to terminate environmental burden activities”. For example, polluting companies make promises to the government and society and agree to carry out environmental governance [27]. The conceptual interpretation of the environmental sector of the European Union pointed out that organizations, including enterprises and groups, make environmental commitments to public rights agencies (including governments) through negotiation [3]. Therefore, it can be seen that an EB is a commitment to environmental protection by an enterprise or an organization. The formulation of an EB is based on the negotiation between the enterprise and the government and uses the binding force of the bargaining to improve the environmental quality and environmental goals.

The truncated decision-making of China’s public policy process will inevitably lead to palpable bargaining during implementation [28]. Environmentalists play a major role in the bargaining process [29]. More and more ENGOs are paying greater attention to the gradually opened and friendly institutional channels of participation [30]. The form of EB is related to the potential of environmental conflict. The higher the potential of conflict, the more informal the form of EB. Most of the EB adopted are direct and informal protection actions. With the resolution of conflicts, the potential for conflicts changes from high to low, and the form of EB changes from informal to formal. Conflicts are mostly manifested in cooperation, persuasion, and inducement. The form is mostly direct and formal, such as multi-party negotiation, joint operation, participation in decision-making, etc. Peng broke down the causes and arrangements of gathering occasions brought about by ecological occurrences in light of social obligation viewpoint [31]. They found that according to the viewpoint of venture parties, their own absence of comprehension of social obligation, the absence of straightforwardness and correspondence with partners, and the less contribution of partners were significant reasons for gathering occasions.

Many scholars have discussed the crucial role that ENGOs have played in the implementation of nature protection policies across European member states. However, there are important differences in the opportunity structures among new and old member states that influence how ENGOs can act and undertake activities [32]. Other scholars have specifically studied the ways for foreign ENGOs to participate in environmental governance [33,34,35], such as promoting legislation, cooperating with the media, and relying on the government. However, most of the research is focused on a single individual, and it is rarely related to the research of multiple subjects composed of government, ENGOs, enterprises, and the public. Only a few articles directly relate to the study the role of ENGOs, but the perspective of EB that interact with the government, enterprises and the public is still lacking. In fact, the protection of the environment is unarguably most important [36].

Many scholars have applied the stakeholder theory to discuss environmental protection issues. However, this theory focuses more on the relationship and roles among stakeholders, which is lacking stakeholder bargaining factors, stage and strategy. Therefore, the EB theory should be combined with the stakeholder theory in order to explore the influencing factors roles and strategies of different stakeholders in different bargaining stages. A conceptual model of ENGO’s participation in environmental protocols was constructed based on the EB and stakeholder theories. This study also compares the role of ENGOs in the Nujiang Dam and Kunming PX incidents, as well as the influencing factors of the role in environmental agreements. The conceptual model is an important innovation of this research, which will provide a groundbreaking theory and research model for the relevant studies.

2. Materials and Methods

2.1. Study Area

China is home to the largest number of dams in the world. Southwest China is in the midst of a dam-building boom—a number of new dams are sparking domestic opposition from grassroots groups and increasing tensions on international rivers such as the Lancang [37]. Yunnan Province is located in Southwest China, and is known for its biodiversity and social assortment. Fifty-four out of the fifty-six officially recognized ethnic minorities in China are in Yunnan Province. Yunnan is rich in water resources, with six huge stream bowls. The hydropower capacity of Yunnan is essentially amassed in the three critical stream bowls of Jinsha River in the upper Yangtze River. Nujiang River flows from the northwest region toward the southeast region. It is generally called the Grand Canyon of the East because of its splendid view containing snow-covered mountains, frigid masses, steep ravines, and solid streams stacked with rapids and raised glades. Before the dam project, the Nujiang River was one of the most beautiful natural streams.

Yunnan Province is also a major petrochemical and energy province in China. The industrial structure of Yunnan’s chemical industry is unbalanced and the resource characteristics of “rich coal and lean oil” are obvious. The inorganic chemical industry, such as the fertilizer and phosphorus chemical industry, has a high concentration and proportion structure. Enterprises, petrochemical, and downstream extension industries are small in scale, and low in proportion. After years of development, Yunnan’s ecological environment has been severely damaged. The deep-seated contradictions in the industry, such as the low degree of resource utilization and the weak ability to resist risks, have become increasingly prominent.

This study mainly selects two cities, Kunming and Nujiang, that are located in southwest China’s Yunnan Province as the case sites (Figure 1). The two case sites were chosen because they are both located in Yunnan and have similar geographical environments, but the outcomes of the EB events are quite different. It is extremely important for this study to analyse the reasons for the differences in their EB through case comparison.

2.2. The Role of ENGOs

Greenspan, et al. [38] have identified four roles for an ENGO, which are entrepreneur, service-provider, enabler, and partner. They then offered an empirical illustration of the typology using eight case studies across the globe and discuss how the four NGO roles might be associated with outcomes of participatory processes. In this study, ENGOs play the role of watchdog, information provider and pressure group in the process of EB. The changes of the role of ENGOs are related to the EB power, EB stage, and EB strategy. A conceptual framework was created by merging the EB and stakeholder theories as shown in Figure 2 to analyse the roles of ENGOs in China. Based on the above theoretical analysis of environmental agreements, this study tested three research hypotheses (

Table 1. Research Hypothesis.

Hypotheses	Definition
H1	Media, human resources, the relationship network, funding, openness, scale, knowledge and experience have significance influence on the role of ENGOs in the Kunming PX Incident.
H2	Media, human resources, the relationship network, funding, openness, scale, knowledge, and experience have significance influence on the role of ENGOs in the Nujiang Dam Incident.
H3	In Kunming PX Incident and Nujiang Dam Incident, there are differences in the order of the importance of media, human resources, the relationship network, funding, open-ness, scale, knowledge and experience to the role of ENGOs

) based on the logical relationship between the factors that influence the role of ENGOs in eight environmental agreements. Three potential exogenous variables set for the study were the role of ENGOs, influence factors of ENGOs, and elements of environmental agreement. Each potential variable is indirectly represented by several observed variables. The role of ENGOs corresponded to three observation variables such as government, watchdog and public. The factors influencing the role of ENGOs are media, human resources, the relationship network, funding, openness, scale, knowledge, and experience (Table 1). Meanwhile, the three observed variables corresponding to environmental agreement elements are EB power, EB strategies, and EB outcomes.

2.3. Structural Equation Model

Structural equation model [39], a technique that is specifically designed to reveal causal relationships between variables, and allows to include hypothetical causal factors, was adopted in this study. Krishnakumar and Ballon [40] mentioned that caution is required when applying SEM, in particular, to emphasize the need for substantive knowledge to drive modeling, exploration, and interpretation of results [41]. AMOS 22.0 software was used to compare and test the fitting degree between the model of sample data and the theoretical hypothesis model and judge the fitting degree results. If the fitting degree meets the standard requirements, the subsequent steps are carried out. If the fit degree does not meet the standard requirements, the model will be modified. The path coefficient test of the modified SEM is carried out, and the utility values of each potential variable are calculated to draw the main conclusions of this study. SEM can be used to study the correlation between multiple variables. Factor analysis and path analysis are the origin of SEM.

SEM, as a widely used method for calculating covariance of variables, scientifically explains the path influence relationship between complex multi-variables. A complete SEM includes two parts: measurement model and structural model. The measurement model is used to describe the relationship between latent variables and their corresponding observed variables. When the measurement model contains multiple latent variables, the causal relationship is used to replace the correlation between the latent variables. The structural model is used to analyse how exogenous latent variables affect the endogenous latent variables and how some endogenous latent variables affect another [42]. The measurement model can be represented by Equations (1) and (2), and the structural model can be represented by Equation (3).

x = Λxξ + δ

(1)

y = Λyξ + ε

(2)

η = Bη + Γξ + ζ

(3)

In Equation (1), x is the observed variable of the external latent variable; Λx is the factor loading matrix, which represents the relationship between the external latent dependent variable and its measured variable x; δ is the error variable that cannot be explained to x; y is the internal latent variable. Observed variable; Λy is the factor loading matrix, representing the relationship between the internal dependent variable and its measurement variable y; ε is the error variable that cannot be explained by the corresponding y. In Formula (3), B is the path influence coefficient matrix between internal latent variables; Γ is the path influence coefficient matrix between external latent variables; ζ is the residual matrix that cannot be explained by the structural equation.

The specific analysis steps are as follows:

(1)

Definition of variables: According to the selected indicators, the concept of each variable is given, and each variable is accurately defined.

(2)

Propose corresponding hypotheses between variables: Based on the theoretical analysis of the indicators, hypotheses are proposed, that is, the causal relationship between the indicators.

(3)

Build a model: According to the assumption proposed in step (2), connect the influence paths to build a model.

(4)

Model estimation and verification: Use the structural equation model verification software to input the constructed theoretical model for fitting, verification, and evaluation, and determine the final evaluation model.

Through the test and evaluation of the hypothesis model, the weights of each index in the model were calculated and determined. The SEM was constructed by combining the scores of each index in the scale. Starting from the eight assumed factors, the indicators affecting ENGO factors are analysed. Firstly, carrying out a quantification analysis of the indicators. According to the Likert 5-level scale method, the scale was compiled, the network research was conducted, and the indicators were quantified. Secondly, SPSS+AMOS data analysis software was used to analyse the questionnaire data. SPSS 26.0 software was used to test the recovered effective scale. The reliability and reliability test and factor analysis of the scale were carried out to determine the final evaluation index system. Then, the AMOS21.0 software was used to test the built model to judge whether the hypothesis is true. According to the path coefficients in the model, the interaction and causal relationship among the factors were determined. Establishment of an evaluation index system and defining the selected influencing factors were conducted through theoretical analysis, determining the causal relationship between the factors, and putting forward hypotheses. Build hypothesis-based theoretical models and structural equations were used to assess factors affecting ENGO elements.

2.4. Questionnaire Survey and Reliability Test

The ENGOs of the “Kunming PX incident” and the “Nujiang dam incident” were selected as the survey objects. A total of 350 questionnaires were distributed to ENGOs that participated in both incidents using a simple random sampling method, of which 272 were valid questionnaires, and the recovery rate was 77.72%. There were 107 male respondents (39.34%) and 165 female respondents (60.66%). The respondents were mainly undergraduate students, with a total of 190 respondents (69.85%), 61 graduate students (22.43%), and 21 respondents with a master’s degree or above (7.72%). The questionnaire is divided into four main parts: the basic situation of the sample, the three stages of the EB, the role of ENGOs, and the factors that influence the role of ENGOs (Supplementary Materials).

Reliability analysis is a test of the stability and consistency of measurement results. In order to ensure the accuracy of measurement results, it is necessary to conduct reliability analysis on the valid data in the questionnaire before analysis. Cronbach’s alpha coefficient is usually used for reliability analysis in social science research. If the reliability coefficient is above 0.9, it means that the reliability is excellent, 0.8–0.9 means good, 0.7–0.8 means acceptable, 0.6–0.7 means questionable, 0.5–0.6 means poor while lower than 0.5 shows unacceptable and needs to be revised. The Cronbach’s alpha coefficient value of the questionnaire form that consists of 31 questions was 0.8, indicating that the questionnaire form is valid and reliable for the actual survey. The reliability coefficient of the items of the scale is relatively high, and the data of this survey are reliable.

2.5. Questionnaire Validity Test

Validity refers to the degree to which the required behavioural characteristics can be accurately measured through a test or scale tool. The Bartlett sphericity test value was used to test whether the correlation coefficient between items is significant. The smaller the significance level of the Bartlett sphericity test (p < 0.05), the more likely there is a meaningful relationship between the original variables. The criteria for whether it is suitable for factor analysis are: greater than 0.9, very suitable; 0.7–0.9, suitable; 0.6–0.7, more suitable; 0.6–0.5 is not suitable; below 0.5, give up. If the significance is less than 0.05, it indicates that each item is suitable for factor analysis. Additionally, the Kaiser-Meyer-Olkin (KMO) value was also used to compare the simple correlation and partial correlation coefficients between items, therefore, the value is between 0 and 1.

Factor analysis was used to analyse the survey results. First, the Bartlett test method was used to obtain that the contribution rate of different items to the factor was greater than 50%, and there was no overlap between the items, indicating that the questionnaire had good validity and was very suitable for factor analysis. The validity is now tested by KMO and Bartlett’s sphericity test. The validity test is used to verify whether the variables are independent of each other through the sphericity test, where the KMO value is 0.816, greater than 0.7, at a 95% confidence level. The validity of the data is considered to be good and can be used for subsequent data analysis as shown in

Table 2. KMO and Bartlett’s test.

	KMO Sampling Suitability Quantity	0.816
Bartlett’s sphericity test	Approximate chi-square	1155.182
	Degrees of freedom	190
	Significance	0.000

.

3. Results

3.1. Environmental Bargaining Model Construction

The covariance solution of the model could be further solved; model indicator variables are shown in

Table 3. Model indicator variables.

Question Number		Nujiang Dam Incident		Kunming PX Incident
		Mean Value	Standard Deviation	Mean Value	Standard Deviation
14.1	High Conflict Stage: Protest	2.094	0.811	2.069	0.751
14.2	Transition Phase: Communication Response	3.150	1.202	2.172	1.169
14.3	Low Conflict Stage: Mutual Negotiation	2.984	1.134	2.807	1.192
17.1	Watchdog (government function supervisor)	3.354	1.185	3.228	1.262
17.2	Government Decision-making Influencer	1.906	1.003	1.772	0.896
17.3	EB Partner	3.283	1.362	2.993	1.233
17.4	Information Provider	2.031	1.408	1.890	1.214
17.5	Public Interest Spokesperson	3.039	1.231	3.028	1.247
17.6	Pressure Groups	3.087	1.209	2.814	1.184
18.1	Experience	2.630	1.153	2.345	1.175
18.2	Human Resources	3.354	1.205	3.076	1.329
18.3	Scale	3.488	1.188	3.228	1.200
18.4	Relation Network	2.992	1.324	2.772	1.268
18.5	Knowledge	3.386	1.209	3.159	1.257
18.6	Media	3.189	1.187	2.779	1.272
18.7	Degree of Openness	2.614	1.169	2.400	1.169
18.8	Funding	3.236	1.294	3.000	1.269
20.1	Bargaining Power	2.921	1.319	2.745	1.306
20.2	Bargaining Strategy	3.276	1.219	2.834	1.318
20.3	Bargaining Outcome	3.354	1.300	3.048	1.260

. The obtained model path results need to be judged for its rationality, and the model adaptation parameters need to be estimated. The general model fitness test indicators and requirements are the ratio of chi-square to degrees of freedom (CMIN/DF < 3), and the average probability root error of square coefficient (RMSEA < 0.08), non-norm fit index (TLI > 0.9), fit index (CFI > 0.9), fit index (GFI > 0.80), canonical fit index (NFI > 0.9) and correction Fit index (IFI > 0.9). After preliminary calculation, the results of fitting parameters are shown in

Table 4. Estimated adaptation parameters of the model.

Fit Metrics	Adaptation Range	Measurements	Does It Meet the Standard
CMIN/DF	<3	2.864	acceptable
RMSEA	<0.08	0.074	acceptable
GFI	>0.8	0.831	acceptable
NFI	>0.9	0.921	acceptable
IFI	>0.9	0.902	acceptable
TLI	>0.9	0.915	acceptable
CFI	>0.9	0.9001	acceptable

CMIN/DF = chi-square/degree of freedom; RMSEA = root mean square error of approximation; GFI = goodness-of-fit index; NFI = normed-fit index; IFI = incremental-fit index; TLI = tucker-lewis index; CFI = comparative-fit index.

., and all the fitting indexes of the equation structure model in this study are acceptable. Therefore, it can be considered that the model has good adaptability, there is a reasonable covariance relationship between variables, and no model correction is required, that is, the collected sample information has a reasonable adaptation to the path covariance calculated by the model. It shows that the model constructed in this study has a reasonable explanation and provides sufficient support for further model path calculation.

3.2. Analysis of Model Results

The structural equation was constructed by sub-sample to test the robustness of the model (

Table 5. The robustness test of the Kunming PX incident (Note: *** = p < 0.001).

Kunming PX Incident					Std Estimate	P	Std Estimate	P1	Std Estimate	P2
@171	watchdog	<-	F1	The role of ENGO	0.131	0.224	0.092	0.545	0.139	0.351
@171	watchdog	<-	@172	Government	0.030	0.744	0.015	0.901	0.042	0.749
@171	watchdog	<-	@173	Enterprises	0.054	0.530	0.116	0.328	0.050	0.694
@171	watchdog	<-	@175	Public	0.170	0.068	0.205	0.122	0.178	0.167
@174	Information officer	<-	F1	The role of ENGO	0.504	***	0.535	***	0.460	***
@176	Pressure group	<-	F1	The role of ENGO	0.581	***	0.561	***	0.586	***
@201	EB power	<-	@171	watchdog	0.112	0.196	0.074	0.510	0.146	0.224
@141	High conflict	<-	@176	Pressure group	0.022	0.789	0.137	0.228	0.103	0.395
@142	Transition	<-	@176	Pressure group	0.134	0.105	0.184	0.102	0.081	0.506
@143	Low conflict	<-	@176	Pressure group	0.087	0.292	0.137	0.228	0.056	0.646
@202	EB strategies	<-	@174	Information officer	0.192	0.015	0.313	0.003	0.085	0.466
@202	EB strategies	<-	@201	EB power	0.302	***	0.488	***	0.365	0.001
@202	EB strategies	<-	@176	Pressure group	0.214	0.589	−0.030	0.768	0.134	0.246
@186	media	<-	F1	The role of ENGO	0.702		0.702		0.740
@185	knowledge	<-	F1	The role of ENGO	0.277	0.004	0.216	0.563	0.232	***
@184	Relationship network	<-	F1	The role of ENGO	0.632	***	0.612	***	0.686	***
@183	scale	<-	F1	The role of ENGO	0.347	***	0.299	0.025	0.351	0.003
@182	Human resources	<-	F1	The role of ENGO	0.660	***	0.704	***	0.689	***
@181	experience	<-	F1	The role of ENGO	0.204	0.032	0.212	0.112	0.211	0.122
@187	openness	<-	F1	The role of ENGO	0.373	***	0.334	0.014	0.366	0.008
@188	funding	<-	F1	The role of ENGO	0.540	***	0.591	***	0.512	***
@203	EB outcomes	<-	@141	High conflict	0.122	0.739	0.015	0.883	0.065	0.516
@203	EB outcomes	<-	@142	Transition	0.083	0.101	0.071	0.054	0.049	0.624
@203	EB outcomes	<-	@143	Low conflict	0.031	0.735	0.046	0.363	0.001	0.992
@203	EB outcomes	<-	@202	EB strategies	0.254	***	0.463	***	0.623	***

and

Table 6. The robustness test results of the Nujiang Dam incident (Note: *** = p < 0.001).

Nujiang Dam Incident					Std Estimate	P	Std Estimate	P1	Std Estimate	P2
@171	watchdog	<-	F1	The role of ENGO	0.187	0.166	0.196	0.293	0.080	0.689
@171	watchdog	<-	@172	Government	0.181	0.142	0.257	0.048	−0.016	0.910
@171	watchdog	<-	@173	Enterprises	0.092	0.915	0.103	0.425	0.136	0.310
@171	watchdog	<-	@175	Public	0.223	0.720	0.027	0.867	0.046	0.791
@174	Information officer	<-	F1	The role of ENGO	0.610	***	0.664	***	0.519	0.003
@176	Pressure group	<-	F1	The role of ENGO	0.531	***	0.603	***	0.444	0.009
@201	EB power	<-	@171	watchdog	0.203	0.316	0.150	0.213	0.029	0.826
@141	High conflict	<-	@176	Pressure group	0.137	0.120	0.045	0.715	0.246	0.053
@142	Transition	<-	@176	Pressure group	0.024	0.789	0.030	0.806	0.016	0.902
@143	Low conflict	<-	@176	Pressure group	0.011	0.905	0.043	0.724	0.045	0.733
@202	EB strategies	<-	@174	Information officer	0.173	0.447	−0.132	0.298	0.197	0.103
@202	EB strategies	<-	@201	EB power	0.562	***	0.409	***	0.398	***
@202	EB strategies	<-	@176	Pressure group	0.224	0.235	0.212	0.092	0.105	0.380
@186	media	<-	F1	The role of ENGO	0.657		0.649		0.633
@185	knowledge	<-	F1	The role of ENGO	0.443	***	0.425	***	0.416	0.198
@184	Relationship network	<-	F1	The role of ENGO	0.666	***	0.674	***	0.701	***
@183	scale	<-	F1	The role of ENGO	0.255	0.011	0.264	0.018	0.232	0.129
@182	Human resources	<-	F1	The role of ENGO	0.618	***	0.605	***	0.615	0.003
@181	experience	<-	F1	The role of ENGO	0.211	0.036	0.206	0.187	0.220	0.120
@187	openness	<-	F1	The role of ENGO	0.477	***	0.483	***	0.494	0.004
@188	funding	<-	F1	The role of ENGO	0.596	***	0.608	***	0.540	0.002
@203	EB outcomes	<-	@141	High conflict	0.083	0.919	0.145	0.214	0.136	0.228
@203	EB outcomes	<-	@142	Transition	0.113	0.030	0.225	0.055	0.174	0.171
@203	EB outcomes	<-	@143	Low conflict	0.165	0.063	0.164	0.171	0.103	0.422
@203	EB outcomes	<-	@202	EB strategies	0.658	***	0.366	***	0.471	***

). The data of the Kunming PX and Nujiang Dam incidents were randomly divided into two parts, and the SEM results of each part were calculated. Comparing the path coefficients of Kunming P1, Kunming P2, and Kunming, the model results were basically the same. The results of Nujiang P1, Nujiang P2, and Nujiang are also basically the same. Therefore, it can be considered that the structural model is a better model, and the results have certain interpretability.

3.2.1. Latent Variable Path Analysis

The calculation principle of the maximum likelihood method was used to calculate the sum of each latent variable. Each latent variable corresponds to the path coefficient among the observed variables, and the operation results are standardized. The path analysis results of the model latent variables obtained from the operation and the analysis results of the standardized path parameters are shown in

Table 7. Hypothesis test results of the Kunming PX incident (Note: *** = p < 0.001).

Kunming PX Incident					Std Estimate	S.E.	C.R.	P
@171	watchdog	<-	F1	The role of ENGO	0.131	0.151	1.215	0.224
@171	watchdog	<-	@172	Government	0.03	0.123	−0.327	0.744
@171	watchdog	<-	@173	Enterprises	0.054	0.087	0.628	0.53
@171	watchdog	<-	@175	Public	0.17	0.093	1.826	0.068
@174	Information officer	<-	F1	The role of ENGO	0.504	0.134	5.098	***
@176	Pressure group	<-	F1	The role of ENGO	0.581	0.139	5.563	***
@201	EB power	<-	@171	watchdog	0.112	0.086	1.292	0.196
@141	High conflict	<-	@176	Pressure group	0.022	0.053	0.268	0.789
@142	Transition	<-	@176	Pressure group	0.134	0.081	1.619	0.105
@143	Low conflict	<-	@176	Pressure group	0.087	0.084	1.054	0.292
@202	EB strategies	<-	@174	Information officer	0.192	0.084	2.435	0.015
@202	EB strategies	<-	@201	EB power	0.302	0.074	5.83	***
@202	EB strategies	<-	@176	Pressure group	0.214	0.084	0.541	0.589
@186	media	<-	F1	The role of ENGO	0.702
@185	knowledge	<-	F1	The role of ENGO	0.277	0.137	2.847	0.004
@184	Relationship network	<-	F1	The role of ENGO	0.632	0.136	6.587	***
@183	scale	<-	F1	The role of ENGO	0.347	0.131	3.558	***
@182	Human resources	<-	F1	The role of ENGO	0.66	0.151	6.528	***
@181	experience	<-	F1	The role of ENGO	0.204	0.125	2.148	0.032
@187	openness	<-	F1	The role of ENGO	0.373	0.128	3.577	***
@188	funding	<-	F1	The role of ENGO	0.54	0.134	5.721	***
@203	EB outcomes	<-	@141	High conflict	0.122	0.116	0.333	0.739
@203	EB outcomes	<-	@142	Transition	0.083	0.077	1.641	0.101
@203	EB outcomes	<-	@143	Low conflict	0.031	0.076	−0.338	0.735
@203	EB outcomes	<-	@202	EB strategies	0.254	0.066	8.123	***

. When α is 0.05, the p values corresponding to the eight influencing factors of NGO that are all less than 0.05, so it is reasonable to think that these eight influencing factors are all statistically significant.

Table 7 shows that the influence coefficients of each latent variable path are significant. From the results of the hypothesis test of the Kunming PX incident model, eight influencing factors of ENGO are ranked according to the size of the standardized influence path coefficients, and the order of influence degree from large to small: Media (Std = 0.702), Human Resources (Std = 0.66), Relationship Network (Std = 0.632), Funding (Std = 0.54), Openness (Std = 0.373), Scale (Std = 0.347), Professional (Std = 0.277) and Experience (Std = 0.204). From the results of the hypothesis test of the Nujiang Dam incident model (

Table 8. Hypothesis test results of the Nujiang dam incident (Note: *** = p < 0.001).

Nujiang Dam Incident					Std Estimate	S.E.	C.R.	P
@171	watchdog	<-	F1	The role of ENGO	0.187	0.205	1.387	0.166
@171	watchdog	<-	@172	Government	0.181	0.112	1.469	0.142
@171	watchdog	<-	@173	Enterprises	0.092	0.08	−0.107	0.915
@171	watchdog	<-	@175	Public	0.223	0.109	−0.359	0.72
@174	Information officer	<-	F1	The role of ENGO	0.61	0.197	5.596	***
@176	Pressure group	<-	F1	The role of ENGO	0.531	0.166	4.948	***
@201	EB power	<-	@171	watchdog	0.203	0.099	1.004	0.316
@141	High conflict	<-	@176	Pressure group	0.137	0.059	1.556	0.12
@142	Transition	<-	@176	Pressure group	0.024	0.089	0.267	0.789
@143	Low conflict	<-	@176	Pressure group	0.011	0.084	−0.119	0.905
@202	EB strategies	<-	@174	Information officer	0.173	0.076	0.761	0.447
@202	EB strategies	<-	@201	EB power	0.562	0.075	4.798	***
@202	EB strategies	<-	@176	Pressure group	0.224	0.087	1.187	0.235
@186	media	<-	F1	The role of ENGO	0.657
@185	knowledge	<-	F1	The role of ENGO	0.443	0.164	4.19	***
@184	Relationship network	<-	F1	The role of ENGO	0.666	0.189	5.976	***
@183	scale	<-	F1	The role of ENGO	0.255	0.152	2.554	0.011
@182	Human resources	<-	F1	The role of ENGO	0.618	0.169	5.663	***
@181	experience	<-	F1	The role of ENGO	0.211	0.149	2.097	0.036
@187	openness	<-	F1	The role of ENGO	0.477	0.159	4.483	***
@188	funding	<-	F1	The role of ENGO	0.596	0.178	5.539	***
@203	EB outcomes	<-	@141	High conflict	0.083	0.128	−0.101	0.919
@203	EB outcomes	<-	@142	Transition	0.113	0.094	−2.169	0.03
@203	EB outcomes	<-	@143	Low conflict	0.165	0.1	1.86	0.063
@203	EB outcomes	<-	@202	EB strategies	0.658	0.085	5.274	***

), it can be seen that the eight major influencing factors of the Nujiang event are ranked according to the size of the standardized influencing path coefficients: Relationship network (Std = 0.666), Media (Std = 0.657), Human resources (Std = 0.618), Funding (Std = 0.596), Openness (Std = 0.477), Professional (Std = 0.443), Scale (Std = 0.255) and Experience (Std = 0.211).

ENGOs have insufficient practical experience in the EB participation for both incidents, which is related to the background of the era when the development of Chinese civil society was relatively slow and the NGO Development Bureau started late. The role of ENGOs in decision-making is becoming more and more important. According to data analysis, the small size of NGOs and weak practical experience have only little influence on the EB. The standardized influence path coefficient shows the media, human resources, and relationship network are three major factors influencing the role of ENGOs in the “Kunming PX Incident” and “Nujiang Dam Incident”, but relationship networks had a greater influence in the latter event, which may have contributed to the successful EB.

The variance inflation factor (VIF) is a measure of the severity of multiple collinearity in a multiple linear regression model. Multi-collinearity means that there is a linear correlation between independent variables, that is, one independent variable can be a linear combination of other one or several independent variables. The most commonly used inspection method is VIF, and the calculation formula is:

$$\mathrm{VIF}=\frac{1}{1-{\mathrm{R}}_{\mathrm{i}}^2}$$

The value of VIF is greater than 1. The closer the VIF value is to 1, the lighter the multi-collinearity, and vice versa. Usually, 10 is used as the judgment boundary. When VIF < 10, there is no multi-collinearity; when 10 < VIF < 100, there is strong multi-collinearity; when VIF > =100, there is severe multi-collinearity. For watchdog (17.1), bargaining strategy (20.2), and bargaining power (20.3) multiple independent variables in the structural equation indicates a strong multi-collinearity.

Table 9. The variance inflation factor values.

Nujiang Event		Collinearity Statistics
Dependent Variable	Independent Variable	Tolerance	VIF
17.1 (watchdog)	17.2	0.868	1.152
	17.3	0.907	1.102
	17.5	0.638	1.568
	F1	0.612	1.635
20.2 (bargaining strategy)	17.4	0.841	1.190
	17.6	0.862	1.160
	20.1	0.972	1.029
20.3 (bargaining power)	14.1	0.962	1.040
	14.2	0.823	1.216
	14.3	0.810	1.234
	20.2	0.972	1.029
Kunming		Collinearity Statistics
Dependent Variable	Independent Variable	Tolerance	VIF
17.1 (watchdog)	17.2	0.865	1.156
	17.3	0.892	1.121
	17.5	0.821	1.218
	F1	0.814	1.228
20.2 (bargaining strategy)	17.4	0.856	1.168
	17.6	0.900	1.111
	20.1	0.949	1.054
20.3 (bargaining power)	14.1	0.953	1.049
	14.2	0.895	1.117
	14.3	0.889	1.125
	20.2	0.970	1.031

shows that the independent variables in each model correspond to the VIF values are all less than 10, showing there is no multi-collinearity.

3.2.2. Path Analysis of the Kunming PX Incident

In terms of the role of ENGOs, the path coefficient (0.58) of Pressure group 17.6 in the role of ENGO impact variable of the Kunming PX event model is higher than that of information officer 17.4 (0.50) and watchdog 17.1 (0.13), indicating that the three roles played by ENGOs in EBs are equally influential (Figure 3). Prominent and all positive, the most prominent role is that of the pressure group, followed by the information provider, and finally the watchdog role. In the Kunming PX incident, the path coefficient of watchdog 17.1 is 0.11 among the influencing variables of EB power 20.1, indicating that ENGOs’ role as a government decision-making supervisor has a positive and significant impact on the power of EBs. Among the influencing variables of EB strategies 20.2, the path coefficient of Information officer 17.4 is 0.19, the path coefficient of Pressure group 17.6 is 0.21, and the path coefficient of EB power 20.1 is 0.30, indicating that the most influential factor on EB strategy is agreement power. Among the influencing variables of EB outcomes 20.3, the path coefficient of EB strategies 20.2 is 0.25, the path coefficient of high-conflict 14.1 is 0.12, the path coefficient of transition 14.2 is 0.08, and the path coefficient of low-conflict 14.3 is 0.03. It shows that all three stages of EB have a positive and significant impact on the result of EB, the high conflict stage of the three conflict stages has the greatest impact on the result of EB, but the main influencing factor of EB in the Kunming PX incident is the environmental strategy, and environmental strategy is mainly influenced by environmental power.

3.2.3. Path Analysis of the Nujiang Dam Incident

Figure 4 shows that the most important role of ENGOs in the Nujiang Dam incident is the information provider (0.61), followed by the pressure group (0.53), and the watchdog (0.19). It shows that ENGO plays three roles in the EB of the “Nujiang Dam Incident”. All the role effects were significantly positive. Among the influencing variables of EB power 20.1, the path coefficient of watchdog 17.1 is 0.20, indicating that ENGOs’ role of government decision-making supervisor has a positive and significant impact on the power of EBs. Among the influencing variables of EB strategies 20.2, the path coefficient of EB power 20.1 is 0.56, the path coefficient of information officer 17.4 is 0.17, and the path coefficient of Pressure group 17.6 is 0.22, indicating that the most influential factor on EB strategy is agreement power. Among the influencing variables of EB outcomes 20.3, the path coefficient of EB strategies 20.2 is 0.66, the path coefficient of high-conflict 14.1 is 0.08, the path coefficient of transition 14.2 is 0.11, and the path coefficient of low-conflict 14.3 is 0.17. It shows that the three stages of the EB have a positive and significant impact on the results of the EB. Among the three stages, the low-conflict stage of the EB of the Nujiang event has the greatest impact on the result of the EB. However, the main factor influencing the outcome of the EB in the “Nujiang Dam Incident” is the environmental strategy, which is mainly influenced by the environmental power.

3.3. Comparison Analysis

The relationship network, media, human resources, and capital are all important factors that affect the role of ENGOs in the Kunming PX incident (Figure 5) and the Nujiang dam incident (Figure 6). The importance of media factors in both events is relatively large. ENGOs played three roles: pressure group, information officer, and watchdog, and all three roles had different degrees of influence on the EB of the two events. In terms of EB, the most influential factor of the EB strategy is the agreement power. All three stages of the EB have a positive and significant impact on the EB result, but the main influencing factor of the EB result in the Nujiang dam incident is the environmental strategy, while the EB strategies are mainly influenced by environmental power.

The media was the most important factor affecting the role of ENGO in the Kunming PX incident (Figure 5), while, in the Nujiang dam incident, the relationship network was the most important factor (Figure 6). From the interview of the Kunming PX event, a new media platform, micro-blogging, plays an important role. An ENGO’s employee said that they used the micro-blogging platform to release a post calling for a parade in the first time, which was then actively responded and forwarded by many citizens. The post soon spread widely on the Internet, especially on Weibo to the entire country. In the Nujiang incident, the person in charge of the ENGO that participated in the event was a media person, so his relationship network was very strong. In addition, he was contacted and supported by the leaders of the local environmental protection Bureau. Some of the leaders even provided him the first relevant information. In the Nujiang Dam incident, some university professors, experts and scholars, government officers, and other very powerful stakeholders also participated in the EB, showing a stronger relationship network than the Kunming PX incident. The success of the Nujiang Dam incident was largely depended on the ENGO’s relationship network.

4. Discussion

Solving the problem of environmental pollution which is restricting the sustainable development of China’s economy and society has become an important subject. Private participation is essential, since it is far from enough to rely only on the government’s “top-down” promotion to solve environmental problems. As representatives of civil society, it is particularly important for ENGOs to promote EB “from the bottom up”. Organizing and participating in the environmental movement is one of the important ways for ENGOS to influence environmental agreements. In most cases, ENGOs are allowed to take part in the discussions and to present substantive documents [43]. ENGOs provide suitable and viable carriers for citizens to participate in the environmental movement, which attracts public support and participation after its launch. Individual power is very limited, but the institutional arrangements of ENGOs can bring together the power of many individual citizens to achieve a certain range or universal reciprocity. As an essential stakeholder of environmental resources, the public has become the third force which assists in promoting environmental governance, together with local governments and polluting enterprises [44]. Although the number of environmental campaigns launched by ENGOs in China is lower than in developed countries, the impact of several important environmental campaigns cannot be underestimated. Two ENGOs in Beijing, Green Home and Yunnan Dazhong Watershed, are at the heart of the campaign against the Nu River Dam. The central government temporarily shelved the Nujiang dam project because of the ENGOs’ appeals and interference. The purely civil environmental movement has finally influenced the government’s decision making, which is a qualitative advance for China’s environmental and social development.

The relation network and media play an obvious role in obtaining resources and winning actions for ENGOs in the EB (Figure 5 and Figure 6). ENGOs are important civil actors in societies’ emergency and disaster responses, and they come together on social media to identify prominent issues and coordinate issue responses [45]. Although both media and ENGOs seem to have different organizational backgrounds, particularly the media is a profitable organization whilst ENGOs are more non-profit oriented, when it comes to environmental communication, most of them agreed that they share quite similar roles particularly in informing and educating the public about environmental issues and in conducting research on environment and sustainability matters [46]. ENGOs have taken advantage of resources and connections via high reputation and trust collaboration to form a triadic relationship of checks and balances with governments and associations [47]. The larger the scale of the organization’s network, the more people it can call on and the more resources it can mobilize. At this time, the resources show the power of solidarity. Social networks are not natural and must be constructed through an investment strategy whose main purpose is to stabilize the relationship so that it becomes a reliable resource [48]. Similarly, in ENGOs’ involvement in the protection of the Baltic Sea, scholars found that the key factors for success were the timing of the success of the NGO initiative, the historic platform for Finn-Russia cooperation in the field of clean water, and the gradual establishment of social networks and legitimacy of actors [49].

In the margins of policy failure or acquiescence, China’s bottom-up organizations tended to rely on the elite figures, mobilizing media and various social forces, using various resources from the people to carry out active activities on certain social issues [50]. Journalists are the major ENGO activists in the Nujiang Dam incident, where colleague relationships also occupy a dominant position in the network resources. These factors ensure ENGOs obtain the support of various resources quickly in social events, and finally win the action.

The factors that have little influence on the outcome of social events include the scale, experience, professionalism, and openness of ENGOs. The success of the Nujiang Dam incident was not due to the large scale of ENGOs. On the contrary, there were not many ENGOs at that time, and they were not registered with the government, belonging to grassroots NGOs. Most ENGOs in China are short of experience and professionalism due to their late start. In addition, in the Nujiang Dam incident at the beginning, a large number of volunteers in the society were not mobilized. The public did not know more about ENGOs, so it did not play a big role in the scale, experience, professionalism, and other factors of the organization, but only played a role in consolidating the achievements of the action in the later period.

5. Conclusions and Recommendations

With the acceleration of economic growth and industrialization, the increasing contradiction between energy supply and demand, as well as the overdraft of environmental bearing abilities, have brought severe challenges to global environmental protection and sustainable development [51,52]. Environmental bargaining (EB) is a process of mutual bargaining among stakeholders. This study aims to analyze the influencing factors on the role of ENGOs in EB of the Kunming PX incident and the Nujiang Dam incident in Yunnan, China. Different EB power brings different EB strategies and ultimately leads to different EB outcomes. Basically, ENGOs play three roles in EB: watchdog, information provider, and pressure group. These three roles enable ENGO to communicate and negotiate with stakeholders in EB and promote the phased development of events. In both incidents, media, relationship networks, and human resources are three major influential factors for the role of ENGOs in EB, but their order of influencing factors is different. In the Kunming PX incident, the most influential factor was media, while a relationship network was the most influencing factor for the Nujiang Dam incident.

Establishing a cooperative protocol network, interacting with other participants, and dealing with power and interest relationships are the key points for ENGOs to play a role in EB. Reasonable site selection of polluting enterprises requires the government, enterprises, ENGOs, and people to form a governance system as the main body, instead of relying solely on the EB system with the government as the main body. In this sense, the location of polluting enterprises is not only a matter of geographical location selection, or a complex public decision-making problem. It is a macro-management-based on the theory of location decision-making and EB, emphasizing the role of civil society, especially non-governmental organizations and the public. It is believed that we should not only make use of the management efficiency of the government, but also play a role in public participation, and implement common governance through cooperation, consultation, partnership, and establishing common goals.

ENGOs play an increasingly important role in EB. Simply guiding the public to directly resist the site selection of polluting enterprises is not only detrimental to the success of project relocation, but also easy to bring many negative results, such as violent conflicts. Therefore, ENGOs in the process of participating in the polluting enterprise site selection should guide the participants to the stage of cooperative game as much as possible. As the representative of the public interest, ENGOs have the responsibility to safeguard the public interest. Therefore, ENGOs, as the “spokesman” and “middleman” of public interests, should integrate various social resources to reach “a bargaining with corresponding rights (discourse right, participation right and decision-making right)”, and coordinate the EBs of various stakeholders. This is important to achieve a scientific and reasonable cooperative game mode of guiding polluting enterprises to select sites. ENGOs should build a rational communication platform between the government and the public, effectively integrate the resource advantages of each participant, mobilize the enthusiasm of all stakeholders, and realize the cooperative game of each participant.

This research has both theoretical and practical significance. Western scholars have completed extensive and in-depth research on ENGOs, while China’s research on ENGOs is relatively limited, so this research topic still needs to be explored further. In particular, the research on the role positioning and influencing factors of ENGOs participating in EB, that is, how ENGOs interact with governments, enterprises, and the public in the process of participating in governance of the environment. Therefore, this study started from the theory of EB and constructed an analysis framework for ENGOs’ participation in EB. With the rise and development of ENGOs in China, ENGOs have become an important governance body for effectively solving environmental problems. Its practical activities in environmental governance are not limited to simple bird watching, tree planting, garbage picking, protection of specific species or environmental education work, but also a strong initiative, third-party supervision, policy participation, interest expression, and maintenance of environmental justice. desire. Therefore, this study tries to analyze the role of ENGOs in EB and the similarities and differences of the influencing factors through the comparative analysis of the two cases according to the EB and stakeholder theory, and finally put forward the role optimization strategy of ENGOs. This has a very important practical significance for analyzing the development direction of ENGOs and broadening its effective participation in EB. The key to the role of ENGOs in environmental agreements is to establish a collaborative agreement network, conduct positive interaction between ENGOs and other stakeholders, and properly handle the power and interest relationship between ENGOs and other stakeholders. How to effectively deal with environmental conflicts in project site selection decisions requires the joint participation of ENGOs and multiple stakeholders to establish a positive interaction relationship of “government-ENGOs-public”, so that the government and ENGOs can effectively cooperate to guide environmental agreements, alleviate environmental conflicts between the public, the government, and polluting enterprises, and promote local sustainable development.

Secondly, ENGOs should make full use of the power of various media, especially new media such as Weibo and wechat, to release relevant information in a real and effective way, so that the public can have a timely and accurate understanding of the changes of their surrounding natural environment, participate in environmental agreements in a timely manner at the early stage of the introduction of pollution projects, and increase the role of prior supervision. Conflicts between polluters and the public should be avoided. Strengthening information exchange, ensuring the smooth flow of environmental information, and protecting the legitimate rights and interests of the public are important actions to be taken by ENGOs in the future. For example, ENGOs should establish internet, telephone, SMS and other channels, and timely disclose the government’s new decision-making trend information and public participation information, so as to effectively play the role of a bridge between the government and the public.

The study is of very significant importance, as there is no comprehensive study conducted in the past in the comparison roles of NGO in EB between “PX event” and “Nujiang event” in Yunnan, China. Therefore, this research study fills this geographical gap. This study presents the current situation, background, and the role that ENGOs have been playing in EB. This study provides a sufficient basis for improvement to government, ENGOs and other stakeholder to recheck their performance and make necessary improvements. The main contribution of this study is to construct a new model based on the environmental protocol and stakeholder theory by combining the three modules of environmental protocol, ENGO role, and ENGO influencing factors using quantitative methods. The research results of ENGOs’ participation in EBs are further innovated.

There are some limitations of this study. For various reasons, it is difficult for this study to personally participate in the numerous environmental conflict incidents in recent years, and there is relatively little first-hand information. The data and cases in the thesis are mostly collected from news media reports using questionnaire information as well as using the literature survey method, and there may be discrepancies between some of the data and information and the actual situation. As some of the information and data involve government and local related interests, the other party is reluctant to disclose them, adding to the difficulty of collecting information and data during the research process. The topic of this study is novel, and there are few references available at this stage, making the construction of the theoretical framework difficult. The research on this selected topic is still in its infancy and the use of theory is immature, and will remain a challenging topic in future research.