Conﬂict Resolution between Multi-Level Government and Farmers in Land Expropriation Based on Institutional Credibility Theory: Empirical Evidence from Shandong Province, China

: Land expropriation has always been a hot spot of social conﬂicts. The land expropriation policy of Merging Villages and Living Together (MVLT) in rural areas has intensiﬁed conﬂicts due to insufﬁcient ﬁnancial compensation and “demolishing old houses before building new ones”. The current research most simply assesses the degree of conﬂict and the inﬂuencing factors but rarely includes farmers, governments at all levels, the strength of policy tools, and policy perceptions in a uniﬁed quantitative research framework, which is not conducive to conﬂict resolution and policy improvement. This paper adopts the institutional credibility theory, incorporates the policy instruments of higher-level governments, administrative instruments of lower-level governments, and farmers’ credibility of policies into a uniﬁed accounting framework, constructs a conﬂict stress index, evaluates the role of each subject’s characteristics, policy perceptions, and policy instruments in the process of conﬂict generation and resolution, and analyzes the methods of conﬂict resolution from the perspective of different stakeholder conﬂicts. The theoretical analysis framework and the quantitative analysis of the indicators are veriﬁed by using a case study of MVLT policy in Shandong Province, China. The results show that the credibility of the policy of “village integration” is inﬂuenced by individual characteristics and varies signiﬁcantly. The administrative means and different combinations of lower-level government are signiﬁcantly related to an increase in farmers’ credibility, which can signiﬁcantly improve the success rate of policy implementation. The effect of administrative means of higher-level government on the credibility of farmers is limited. The highest value of the conﬂict index was observed when the administrative instruments reached the maximum value without a marginal increase in farmers’ credibility. Based on the quantitative evaluation of conﬂict generation and resolution mechanisms, recommendations for policy implementation and improvement were made.


Introduction
Land expropriation in China (hereinafter referred to as land expropriation) is an important realistic path for rapidly promoting urbanization. From 2003 to 2017, the government expropriated a total of 5,555,297 thousand hectares of land at an average of 370,353 thousand hectares per year [1]. This results in a loss of farmers' land use and usufruct rights, and the resulting insufficient compensation is the primary reason for land expropriation conflicts [2][3][4]. Cai et al. (2020) [5] pointed out that there is an enormous deficiency between the compensation for land-lost villagers and the market value of land. Therefore, insufficient land expropriation compensation has been criticized as a deprivation for landlost farmers [6,7]. It is also the most important source of social unrest and the significant number of conflicts caused by land expropriation in China's rural and surrounding urban areas [8][9][10][11]. Over the past decade, more than 65% of Chinese farmers' petitions are due to land issues, of which 73.2% are related to land expropriation [12]. These latter conflicts have become a key issue affecting rural stability in China [13].
Although it is a type of land expropriation, the policy of Merging Villages and Living Together (MVLT) is different from an ordinary one that is government led. The implementation of the MVLT is related to the policy of " linking newly added cropland quotas with the amount of land used for urban and rural construction (LID project)" issued by the Ministry of Natural Resources [14,15]. In order to save land resources, the policy allows local governments to increase the index of urban construction land by reducing the amount of rural construction land in the same area. The specific approach of the government-led policy for MVLT in rural areas is to demolish peasant households in several neighboring rural areas. The homesteads will be recovered, and new rural communities will be composed of buildings to resettle the demolished peasant households. This will achieve land reclamation and increase the area of cultivated land in exchange for an increase in urban construction land indicators. The original intention of the LID project was to protect China's total control target of 1.8 billion mu of arable land, which can change dynamically in terms of spatial distribution but maintains a balance in terms of the total amount. Implementing the MVLT policy can result in residents living centrally, thus effectively reducing rural construction land that will gain urban construction indicators instead.
In addition to insufficient compensation, the most significant characteristic of the MVLT policy is "demolishing houses first and then building resettlement houses". Before the new houses are constructed, the old houses have to be demolished, and the homestead rights of farmers are, therefore, infringed [16]. The farmer must build temporary housing on his contracted land before the resettlement housing is completed or some of them "live in no fixed place". Even if they live in resettlement houses, the original form of open living becomes centralized apartment living, which leads to "separation of production and living functions" [17]. This directly causes economic risks for migrant farmers due to the breakdown of livelihood networks [18]. Chuang (2015) [1] believed that many land expropriation farmers have become "homeless without three things", that is, no land, no job, and no basic welfare guarantee [19]. Therefore, compared with the general land expropriation policy, the MVLT policy receives more resistance from farmers, which has led to contradictions and conflicts that are more intense.
There are two completely opposite explanatory paradigms in academia regarding conflicts during land expropriation in China. One is the pattern of peasant rights protections, which suggests that lower governments violate the law and infringe upon the rights of farmers during land expropriation, leading to their resistance and defense of their rights [20][21][22]. The other explanatory paradigm is the interest conflict paradigm, which suggests that there are many vague interest spaces in the process of land expropriation. All interest entities involved in land expropriation engage in intense interest conflicts to maximize their benefits, resulting in land expropriation conflicts, including petitioning and group conflict events [23][24][25][26]. In addition to the above two explanatory paradigms, there is a third explanatory paradigm for land expropriation conflicts. It suggests that the main reason for conflicts is the unreasonable land expropriation system, which is characterized by low compensation, excessive scope, and lack of negotiation, and non-public interest expropriation is the main cause of land expropriation conflicts. This explanatory paradigm can be called the institutional paradigm [27][28][29][30]. These three paradigms analyze the causes and resolution of conflicts in the process of land expropriation from different perspectives and need to be studied from a more comprehensive perspective.
Regarding the degree of conflict in land expropriation, Liu Haoran and Liang Fachao (2020) constructed a framework of factors that affect the intensity of land expropriation conflicts and qualitatively analyzed the intensity of 12 land expropriation conflicts [31]. Lin et al. (2018) classified China's land conflicts into four types based on four indicators: direct economic losses caused by conflicts, number of participants, injuries, and deaths [32]. These types are massive land conflicts, serious land conflicts, ordinary land conflicts, and non-land conflicts. Obviously, this quantitative classification is inaccurate and difficult to apply because there is no transitional state between these four types. It is difficult to show the degree of conflict without personnel casualties, and it is impossible to show the interaction between farmers and government in land expropriation based on the scale of conflicts.
As China's most controversial land policy, the MVLT has resulted in numerous conflicts, and the degree of conflict involved in MVLT is far greater than that of regular land expropriation. However, MVLT has not incited sustained and large-scale social conflicts and unrest. Although MVLT has become a profound problem, there is a lack of quantitative analysis of the degree of conflict and its resolution mechanism. Therefore, how to resolve MVLT conflicts is the main focus of this article. This paper will analyze the policy credibility of farmers, the policy tools used by the upper government to implement the policy objectives, and the administrative means used by the lower government to organize farmers. It proposes a theoretical analysis framework for the conflict between farmers and the government that explains how the conflicts were generated and how they can be resolved. The conflict pressure index and the conflict index as well as their calculation methods are constructed, and the changes in these indexes reflect the role of administrative means of lower government and policy tools of upper government.
There are three innovations in this paper: (1) The farmers' behavior will, the upper government's will, and the lower government's administrative means are brought into the theoretical framework of a conflict analysis, which provides a broader vision for conflict generation and solution. (2) The conflict pressure and the conflict itself are distinguished, and the conflict pressure and conflict index are quantitatively measured, which provides a quantitative basis for the analysis of conflict intensity. (3) This paper discusses the mechanism of conflict resolution, analyzes ways to resolve land expropriation conflicts with the change in policy tools of the upper government and administrative means of the lower government. This also explains why there are no continuous and major conflicts leading to endangering social stability in the implementation of the MVLT policy.
This paper is divided into six sections. The first part is the Introduction. The second is the theoretical analysis framework and hypotheses, including the mechanism of how farmers' willingness and government behavior affect the generation of the conflict. The third section comprises an overview of the study area, the acquisition of data, and the introduction of research methods. The fourth section is an analysis of the results, the fifth is a Discussion, and the sixth is the Conclusions.

The Stakeholders in the MVLT Conflict
During the MVLT process, various major stakeholders are involved [33]. Fan et al. (2016) believed that the stakeholders involved in rural land expropriation include expropriated farmers, lower governments, upper governments, rural collective economic organizations, netizens, and the media [34]. Tan and Tu (2009) stated that the stakeholders in land expropriation conflicts mainly include farmers, enterprises, lower governments, and rural collective economic organizations [35].
The stakeholders involved in land expropriation are directly related to the causes of conflicts and the ways to resolve them. The method for determining the main stakeholders in the land expropriation process is to judge whether they have the right to directly dispose of land interests. Farmers have the right to use land, and the economic benefits of land production are the farmers' personal property. However, land expropriation deprives farmers of their land rights, including disposal rights and other related rights, resulting in a significant increase in their living costs. When farmers' own interests are damaged, they often develop a confrontational mentality against the government, which can lead to conflict. Therefore, expropriated farmers are one of the main stakeholders in land expropriation conflicts.
Rural collective organizations are the owners of collective land. However, since the promulgation of the "Opinions on Improving the Separation of Rural Land Ownership, Contract Rights, and Management Rights" 1 by the Central Committee of the Communist Party of China and the State Council in 2016, which established the "separation of three rights" of rural land ownership, contract rights, and management rights, and gave farmers the right to dispose of land contract rights, rural collective organizations can no longer be considered as independent stakeholders.
Netizens are not direct stakeholders in land expropriation and are more concerned about the disadvantaged groups in the process of land expropriation due to psychological identification. Similarly, the media are not a direct stakeholder in land expropriation and obtain attention or satisfaction of professional missions from the process. Therefore, neither netizens nor the media are stakeholders in land expropriation.
The government is the producer of land expropriation policies. In implementing land expropriation policies, the government can adopt both market and administrative approaches. Generally, the government uses demolition or real estate companies to carry out land expropriation development projects that are urgent and have high land prices. The benefits of this approach are that it can ensure efficiency and avoid direct confrontation with expropriated farmers. However, the downside is obvious: companies are not familiar with the specific situation of expropriated farmers, lack channels for communication with farmers, and administrative capabilities, which often exacerbate land expropriation conflicts and lead to serious conflicts. According to Lin et al., most land conflicts are caused by companies implementing land expropriation [32]. Therefore, to ensure social stability and protect the interests of expropriated farmers, since the promulgation of the "separation of three rights" of rural land in 2016, the frequency and scale of market-oriented land expropriation implemented by upper governments have been significantly reduced.

The Conflict of the MVLT and Farmers' Credibility
The Conflict Analysis Model (CAM) originated from the study of land tenure and forest conflicts [36] and was gradually developed to include a set of different variables [37], notably, timing, source, nature, frequency, intensity, duration, and outcome of conflicts. This was later expanded with additional indicators such as the different actors involved in a conflict [38]. The model is a heuristic tool to which indicators can be added, adjusted, and operationalized according to the needs of the study. Its aim is to approach conflict in a multi-dimensional and temporally and spatially sensitive manner by going through a reiterative process of hermeneutical data interpretation. In effect, it is a flexible, analytical instrument for assessing the relevant variables, rather than a rigid model in which each indicator needs to be present. To date, the CAM has been applied and tested through a variety of studies, such as on urban commons [39], grassland [40], and indigenous land rights [41], and has become a tool to evaluate the conflict of farmland systems [38,42].
The CAM also has certain areas for further refinement: (1) It mostly describes the conflict extent and the results and focuses less on the drivers of the conflict; thus, there is a lack of a basis for taking effective measures to resolve conflicts. (2) For a land conflict that is a result of public policy and public governance, the role of the government as a policy maker and promoting executor is important. What role the government plays in conflict generation and resolution has not been studied. Thus, solutions could be found from the credibility theory to resolve the above problems.
The "credibility thesis" provides a corresponding relationship between institutional credibility and behavioral response. Grabel (1994) pointed out that, when the policy is "credible", a rational economic man will respond to market signals as described by the neoclassical theory [43]. Peter Ho systematically proposed the theory and measurement method (institutional credibility thesis) of institutional "credibility" and defined it as follows: When a certain institution exists and persists, it likely plays a certain role or fulfills a function in society [36,37,41]. Social and economic executors have a certain degree of perception and support for the system, believing that it is credible, will implement it, and it is efficient. Otherwise, the efficiency of the system is relatively low. Farmers' credibility consists of the perception of three dimensions: profit and loss, conflict, and institutional change [40,41,44]. According to the value of the credibility, institutional credibility is divided into five levels: high, medium-high, neutral, medium-low, and low, corresponding to the different levels of policy intervention [45]. Thus far, institutional credibility theory has been widely used in public policy and has successfully explained numerous phenomena [38][39][40][41][45][46][47][48], indicating the wide application of the theory.
The credibility thesis and its underlying theory hold that credibility is endogenous. It is significantly affected by the characteristics of farmers' age, gender, education level, household income, and income sources [45,49,50]. Diniz et al. (2013) [51] found that the five types of livelihood capital of farmers, including human, material, natural, social, and financial resources, have an important impact on farmers' livelihood strategies and are an important driving force for farmers' willingness to participate in land reform. Fan (2022) analyzed the implementation of grassland ecosystem service policies and believed that the livelihood capital and livelihood type of farmers are endogenous to credibility [52]. It is sufficient to observe that the credibility of policies will be different if the characteristics of policy implementers are also different, and the policies' results will vary. There is a causal chain between individual characteristics, credibility, and policy results ( Figure 1). In this context, the following hypothesis for the MVLT policy is proposed:

H1:
There is a significant correlation between the individual characteristics of farmers and the credibility of the MVLT. If the individual characteristics of farmers, such as living place, income level, and social capital, are different then the credibility of the policy is also significantly different.
neoclassical theory [43]. Peter Ho systematically proposed the theory and measurement method (institutional credibility thesis) of institutional "credibility" and defined it as follows: When a certain institution exists and persists, it likely plays a certain role or fulfills a function in society [36,37,41]. Social and economic executors have a certain degree of perception and support for the system, believing that it is credible, will implement it, and it is efficient. Otherwise, the efficiency of the system is relatively low. Farmers' credibility consists of the perception of three dimensions: profit and loss, conflict, and institutional change [40,41,44]. According to the value of the credibility, institutional credibility is divided into five levels: high, medium-high, neutral, medium-low, and low, corresponding to the different levels of policy intervention [45]. Thus far, institutional credibility theory has been widely used in public policy and has successfully explained numerous phenomena [38][39][40][41][45][46][47][48], indicating the wide application of the theory.
The credibility thesis and its underlying theory hold that credibility is endogenous. It is significantly affected by the characteristics of farmers' age, gender, education level, household income, and income sources [45,49,50]. Diniz et al. (2013) [51] found that the five types of livelihood capital of farmers, including human, material, natural, social, and financial resources, have an important impact on farmers' livelihood strategies and are an important driving force for farmers' willingness to participate in land reform. Fan (2022) analyzed the implementation of grassland ecosystem service policies and believed that the livelihood capital and livelihood type of farmers are endogenous to credibility [52]. It is sufficient to observe that the credibility of policies will be different if the characteristics of policy implementers are also different, and the policies' results will vary. There is a causal chain between individual characteristics, credibility, and policy results ( Figure 1). In this context, the following hypothesis for the MVLT policy is proposed:

Government Behavioral Will and Generation of Conflict
Public choice theory holds that the government is also an "economic man" with its own pursuit of interests [53,54]. Therefore, the analysis of the government's willingness should be based on the pursuit of the interests (utility) of land expropriation. Government has a hierarchical structure (Tirole 1986) [55], and it is simplified into two levels for the MVLT, i.e., upper and lower government.

Government Behavioral Will and Generation of Conflict
Public choice theory holds that the government is also an "economic man" with its own pursuit of interests [53,54]. Therefore, the analysis of the government's willingness should be based on the pursuit of the interests (utility) of land expropriation. Government has a hierarchical structure (Tirole 1986) [55], and it is simplified into two levels for the MVLT, i.e., upper and lower government.
The upper government formulates the policy of the MVLT. There are two major driving factors for upper government to do so: one is to develop the local economy and reduce fiscal deficit [5,[56][57][58], and the second is the motivation to promote government officials [59][60][61]. Li and Zhou verified the significant correlation between the promotion of local officials and local economic performance [62]. In addition, with the intensification of social contradictions in recent years, the risk of social stability caused by possible major conflicts has also become an important factor to be considered in the upper government's policy formulation [10,63]. Therefore, it will assess the social stability problems that may be caused by the MVLT before formulating it [64]. When the risk can be controlled, the upper government will depute the lower government to implement the MVLT. In order to Land 2023, 12, 844 6 of 22 ensure its effective implementation, the upper government uses policy tools, including task assignment, inspection and acceptance, incentive measures, and performance assessment, to ensure the lower government's policy implementation [65][66][67].
The lower government is also an "economic man", which is self-interested in the structure of the principal agent in policy implementation [68,69]. Under the restriction of the policy tools of the upper government, lower-government officials implement the MVLT and achieve the policy objectives in their pursuit of political achievements. To achieve this, the lower government will use all administrative resources to complete the policy of the upper government within the prescribed timeframe [70,71]. Administrative means, such as land compensation discretion, administrative penalties, and organizational mobilization, are used to mobilize farmers to implement the MVLT.
Various major stakeholders are involved in the MVLT [33][34][35]. There are three primary actors: upper government, lower government, and peasant households. The process of conflict and interaction among them promotes the implementation of the MVLT but also produces prominent contradictions ( Figure 2). In Figure 2, the upper government entrusts the MVLT to the lower government and, at the same time, urges the lower government to implement it for the policy objectives through policy tools. The lower government acts on behalf of the upper government's policy and mobilizes farmers to participate through administrative instruments. Farmers decide whether to participate in the policy according to their own assessment of the policy's credibility. The conflict pressure in the process of the MVLT comes from the gap between the credibility of the upper government and the farmers, which will lead to conflict under the promotion of the policy tools of the upper government. We use the conflict index to measure the degree of conflict. [59][60][61]. Li and Zhou verified the significant correlation between the promotion of local officials and local economic performance [62]. In addition, with the intensification of social contradictions in recent years, the risk of social stability caused by possible major conflicts has also become an important factor to be considered in the upper government's policy formulation [10,63]. Therefore, it will assess the social stability problems that may be caused by the MVLT before formulating it [64]. When the risk can be controlled, the upper government will depute the lower government to implement the MVLT. In order to ensure its effective implementation, the upper government uses policy tools, including task assignment, inspection and acceptance, incentive measures, and performance assessment, to ensure the lower government's policy implementation [65][66][67].
The lower government is also an "economic man", which is self-interested in the structure of the principal agent in policy implementation [68,69]. Under the restriction of the policy tools of the upper government, lower-government officials implement the MVLT and achieve the policy objectives in their pursuit of political achievements. To achieve this, the lower government will use all administrative resources to complete the policy of the upper government within the prescribed timeframe [70,71]. Administrative means, such as land compensation discretion, administrative penalties, and organizational mobilization, are used to mobilize farmers to implement the MVLT.
Various major stakeholders are involved in the MVLT [33][34][35]. There are three primary actors: upper government, lower government, and peasant households. The process of conflict and interaction among them promotes the implementation of the MVLT but also produces prominent contradictions ( Figure 2). In Figure 2, the upper government entrusts the MVLT to the lower government and, at the same time, urges the lower government to implement it for the policy objectives through policy tools. The lower government acts on behalf of the upper government's policy and mobilizes farmers to participate through administrative instruments. Farmers decide whether to participate in the policy according to their own assessment of the policy's credibility. The conflict pressure in the process of the MVLT comes from the gap between the credibility of the upper government and the farmers, which will lead to conflict under the promotion of the policy tools of the upper government. We use the conflict index to measure the degree of conflict. When the MVLT is promulgated by the upper government, it means that the policy is completely credible. The value of credibility of the upper government is 1 on a value When the MVLT is promulgated by the upper government, it means that the policy is completely credible. The value of credibility of the upper government is 1 on a value range of credibility between 0 and 1. If farmers also believe that the policy is credible, they will actively cooperate with the implementation of the policy [37], and the pressure of conflict will be minimal at this time. On the contrary, if farmers think that the policy is not credible and their willingness to implement the policy is very low, there will be a fierce conflict between the government and farmers in the MVLT process. With a greater gap of credibility, there is greater conflict pressure; the smaller the gap, the lower the conflict pressure. The expression is shown in the formula below: In Formula (1), CFI is the conflict stress index; C g is the government's credibility; and C f is the farmers' credibility.
The high pressure of conflict does not mean that the conflict is strong because it also depends on the intensity of the implementation of policies by the lower government. Its driving force to implement policies comes from the policy tools of the upper government. If the upper government proposed the task of the MVLT to the lower government without corresponding policy tools, that is, without taking any governance measures, that means the policy is "a dead letter". Consequently, the lower government may not act. In this situation, there is no possibility of conflict.
The conflict index is defined as: In Formula (2), CFindex is the conflict index; PTI is the policy tools index.
In the value range of a conflict index between 0 and 1, 1 is absolute conflict for which there is no possibility of reconciliation and land expropriation is either enforced by the government or terminated; 0 means that there is no conflict, and the land requisition is carried out smoothly.
The Pareto 80/20 rule is very important for the revelation of land expropriation conflicts. In studying these conflicts [8,12,13], all of the focus will be on analyzing the significant role of the people who have the least willingness for land expropriation that causes conflicts and rural stability problems. In addition, there will be an examination of Lo's (2021) research on the new poverty problem caused by insufficient compensation for the prohibition of logging in China [72]. The attention to Hu's (2020) [73] research on the violation of voluntary and consent rights of an individual in the "coal-to-gas" project in rural China will be centered on vulnerable groups. Therefore, we need to pay attention not only to the willingness and conflict of all of the land-expropriated farmers but especially to the conflicts caused by the 20% of farmers with the lowest willingness to have their land expropriated. We define the conflict value of the latter group as: In Formula (3), C f ≤ 20% is the arithmetic mean of the credibility of the lower 20% of farmers in the sequence of all land-expropriated farmers from low to high.

Administrative Means of Lower Governments and Conflict Resolution
The biggest feature in the implementation of land expropriation policies is the negotiation and bargaining between government and farmers in the process of the implementation of the land expropriation policy [64,74]. Most of these conflicts are alleviated or resolved through bargaining between the lower government and farmers in various stages, and the land expropriation policy is finally implemented for all or most of the farmers.
From Formula (2), it can be seen that the resolution of the conflict depends on reducing the PTI of the upper government or improving the credibility of farmers. Generally speaking, the PTI of the upper government is not easy to change; thus, changing the credibility of land expropriated farmers becomes key to resolving conflicts and completing land expropriation tasks.
How do farmers determine credibility? According to Peter Ho's credibility measurement [37,45], the credibility of land expropriation policy depends on three secondary indicators, specifically, farmers' perception of profit and loss, conflict, and institutional change ( Figure 3). The perception of profit and loss is an estimation of whether the amount of compensation for demolition can balance the loss. The perception of conflict is the estimation of the contradiction with the local government, other farmers in the same village, and relatives. The perception of institutional change is the prediction of possible policy changes if they insist on refusing demolition. These three indicators of credibility fully reflect the overall perception of farmers with the land expropriation policy and determine their willingness to act. change( Figure 3). The perception of profit and loss is an estimation of whether the amount of compensation for demolition can balance the loss. The perception of conflict is the estimation of the contradiction with the local government, other farmers in the same village, and relatives. The perception of institutional change is the prediction of possible policy changes if they insist on refusing demolition. These three indicators of credibility fully reflect the overall perception of farmers with the land expropriation policy and determine their willingness to act. How do administrative means affect the credibility of farmers? Those used by the lower government primarily include the discretion of compensation, the power of administrative punishment, and the ability of organization and mobilization( Figure 3). It is easily discerned how the administrative means affect the credibility of farmers when we further divide it into three levels of indicators ( Table 1). The separate or combined use of these three administrative means in Table 1 may have an effect on the perception of profit and loss, conflict, and institutional change, thus affecting farmers' credibility. Under the threat of administrative means, the value of this may increase and resolve the conflict. To this end, we propose the following hypothesis: H2: There is a significant correlation between the administrative means of the lower government and an improvement in farmers' credibility, and different administrative means or combinations of them have different effects on the credibility of policies.

Administrative means
Discretion in compensation expenses Adopt the way of increasing or reducing the compensation standard for demolition and removal Authority of administrative penalties (hard means) Use a way of limiting freedom Way that threatens the employment or welfare of family members or relatives Take water disconnection, power cut, and road closure Threat forced demolition Organization and mobilization ability (soft means) Constantly sending people to persuade them Party members or village cadres to take the lead and asked to persuade other farmers How do administrative means affect the credibility of farmers? Those used by the lower government primarily include the discretion of compensation, the power of administrative punishment, and the ability of organization and mobilization ( Figure 3). It is easily discerned how the administrative means affect the credibility of farmers when we further divide it into three levels of indicators ( Table 1). The separate or combined use of these three administrative means in Table 1 may have an effect on the perception of profit and loss, conflict, and institutional change, thus affecting farmers' credibility. Under the threat of administrative means, the value of this may increase and resolve the conflict. To this end, we propose the following hypothesis: H2: There is a significant correlation between the administrative means of the lower government and an improvement in farmers' credibility, and different administrative means or combinations of them have different effects on the credibility of policies. People whose family members work in government and business persuade their family to agree to the demolition.
Ask your relatives to do the persuasion work The administrative means of the lower government can improve farmers' credibility and reduce the pressure of conflict. However, the role of administrative means in resolving conflicts is limited. When all means fail to achieve the goal of improving farmers' credibility, the lower government under the strong restriction of the time node of land expropriation by the upper government will use administrative power to force demolition for the remaining farmers with lower credibility without consent. This type of administrative means goes against the wishes of farmers, which will inevitably lead to fierce conflicts and a rise in conflict pressure. To this end, the authors propose the following hypothesis: H3: Improvements in the credibility of farmers via administrative means is limited. When the lower government using the maximum administrative means also fails to increase the credibility of farmers, the degree of conflict is highest.
Generally speaking, the upper government cannot easily change the MVLT after it is promulgated; thus, the PTI is relatively stable. According to Formula (2), when the PTI is constant, the size of conflict is related to farmers' credibility index. Based on hypothesis H1, H2, and H3, we propose the following ratiocination: Ratiocination 1: Conflict is determined by the individual characteristics of the behavioral agent, and there are different conflict values with various places of residence, income levels, social capital, etc.

Ratiocination 2:
The separate and combined use of different administrative means will effectively resolve the conflict.

Ratiocination 3:
The conflict index reaches its maximum when the administrative means of the lower government are completely used up, and the conflict breaks out.
When the policy credibility gap between the government and farmers reaches its maximum, the greatest conflict pressure occurs, which will inevitably lead to fierce conflict. When information, such as the degree and scale of conflict, in the MVLT is transmitted to the upper government, it will assess whether the conflict seriously endangers social stability. The central government retains "veto power" and the accountability system over the decisions of the lower government that pose a major risk to social stability [10,63]. When the upper government assesses that the continued implementation of the MVLT will greatly affect social stability, the MVLT will be terminated, and the PTI will become zero. According to Formula (2), the conflict index is 0 at this time, and the conflict disappears. It can be seen that the upper government's timely termination of the MVLT is an effective way to resolve conflicts. At the end of the MVLT, the remaining farmers become "nail households" [75,76]. In this changing governance system, the phenomenon of the existence of uncompromising nail households reflects the important role of the social stability mechanism that ultimately controls land expropriation conflicts within a certain range.

Overview of the Study Area
Tancheng County is a county under the jurisdiction of Linyi City, Shandong Province. Tancheng County is located in the low mountains and hills in central and southern Shandong Province and in the hinterland of the Tancang Plain ( Figure 4). Tancheng County is a jurisdiction with 13 towns. At the end of 2020, the total number of households was 314,700, the total population was 1,045,400, the urban population was 308,500, and the rural population was 736,900. It is an agricultural county for which the GDP was CNY 32.094 billion in 2020. The added value of the primary industry was CNY 3.587 billion accounting for 11.2% of GDP; the second was CNY 8.408 billion accounting for 26.2% of GDP; and the third was CNY 20.099 billion accounting for 62.6% of GDP.
In Tancheng County, the villages of Shengli Town, Lizhuang Town, Gangshang Town, Gaofengtou Town, and Miaoshan Town were selected for implementing the LID project with the MVLT in March 2018. Compensation for house demolition shall be implemented in accordance with the "Notice of the Office of the People's Government of Tancheng County on Publishing the Compensation Standards for Ground Attachments and Green Crops in Land expropriation" issued on 6 January 2016. It is publicly announced that the MVLT is the project LID. Prior to its initiation, the local government entrusts the relevant real estate estimation company with measuring and registering the houses and ground attachments of each household and estimates the total compensation. Various compensation amounts will be paid for different housing types and construction standards, and they are approximately equivalent to the replacement construction cost; however, there is no compensation for the homestead area. In order to promote rapid implementation, the county government established a leading group for the work of the LID project. It is headed by the executive deputy head of the Standing Committee of the County Party Committee, with a special office under it that is responsible for the specific activities of farmers' housing demolition and resettlement housing construction as well as inspecting and supervising the project's progress.
Land 2023, 12, x FOR PEER REVIEW 10 Shandong Province and in the hinterland of the Tancang Plain ( Figure 4). Tanch County is a jurisdiction with 13 towns. At the end of 2020, the total number of househ was 314,700, the total population was 1,045,400, the urban population was 308,500, the rural population was 736,900. It is an agricultural county for which the GDP was C 32.094 billion in 2020. The added value of the primary industry was CNY 3.587 bi accounting for 11.2% of GDP; the second was CNY 8.408 billion accounting for 26.2 GDP; and the third was CNY 20.099 billion accounting for 62.6% of GDP. In Tancheng County, the villages of Shengli Town, Lizhuang Town, Gangsh Town, Gaofengtou Town, and Miaoshan Town were selected for implementing the project with the MVLT in March 2018. Compensation for house demolition shall be plemented in accordance with the "Notice of the Office of the People's Governme Tancheng County on Publishing the Compensation Standards for Ground Attachm and Green Crops in Land expropriation" issued on 6 January 2016. It is publicly nounced that the MVLT is the project LID. Prior to its initiation, the local governm entrusts the relevant real estate estimation company with measuring and registering houses and ground attachments of each household and estimates the total compensa Various compensation amounts will be paid for different housing types and construc standards, and they are approximately equivalent to the replacement construction however, there is no compensation for the homestead area. In order to promote rapid plementation, the county government established a leading group for the work of the project. It is headed by the executive deputy head of the Standing Committee o County Party Committee, with a special office under it that is responsible for the spe activities of farmers' housing demolition and resettlement housing construction as we inspecting and supervising the project's progress.

Data Acquisition
In this paper, Q village in Shengli town was selected from five villages in Tanch county as the investigation area. The first author's parents live there and investigated entire process of the implementation of the MVLT as a local and acquired first-hand in its process. Q village is located in the northwest of Tancheng County, ten kilometers away f the county seat and one to two kilometers away from Shengli Town. There are 1215 ho holds in the village, and 351 of them participated in the project, accounting for 28.

Data Acquisition
In this paper, Q village in Shengli town was selected from five villages in Tancheng county as the investigation area. The first author's parents live there and investigated the entire process of the implementation of the MVLT as a local and acquired first-hand data in its process. Q village is located in the northwest of Tancheng County, ten kilometers away from the county seat and one to two kilometers away from Shengli Town. There are 1215 households in the village, and 351 of them participated in the project, accounting for 28.9% of the total number of households in the village. Beginning on 12 April 2018, 15 staff members (all civil servants) in charge of demolition in Shengli Town and 50 village cadres in Q village formed a working group for the LID project. They began to distribute publicity materials, issued compensation standards of demolition to each peasant household, and persuaded them to sign demolition agreements. Since 14 April 2018, our survey team has conducted a full sample follow-up survey on the 351 relocated households, primarily based on a Participatory Rural Appraisal (PRA) [77,78]. We use PRA tools, such as questionnaires, observations, and semi-structured interviews, to finalize our survey. See Table 2 for basic information on the relocated farmers.

Measurement Index of Farmers' Credibility
We refer to the credibility measurement of Peter Ho [37,45] and divide the credibility into three secondary indicators of profit and loss perception, conflict perception, and institutional change perception, according to the specific situation of MVLT. Each secondary indicator is subsequently broken down into three-level indicators.
The indicator settings are described as follows: (1) The value of the profit and loss perception is obtained from the following threelevel indicators: a. Can the compensation fund for land expropriation policy balance profits and losses? Value of loss balance = amount of compensation -increased cost. If the value is >0, it is income; if the value is <0, it is a loss. This is set as follows (single choice): The compensation given by the government is completely sufficient; sufficient; enough; not enough; insufficient.
b. Is there any delay in the gain compensation funds for the land expropriation policy? This is set up as follows (single choice): Immediately, half a year, 1 year, 2 years, and more than 2 years.
(2) The value of the conflict perception is obtained from the following three-level indicators: a. The possible size of the conflict? This is set as follows (single choice): No conflict, agree to land expropriation; minor conflict; moderate conflict; fierce conflicts, petitions, and appeals; serious conflict, vicious fight, and death.
b. In which groups does conflict occur? This is set as follows (multiple choice): Between farmers and local governments; among farmers in the same village; between farmers and their relatives and social relations.
(3) The value of institutional change is obtained from the following three-level indicators: a. Whether the land expropriation policy is consistent? This is set as follows (single choice): No change; resist for a while, the government will raise the compensation standard; after a period, the land expropriation policy will come to an end.
b. If farmers refuse to change, what is the possibility of the government changing its policy? This is set as follows: Impossible, maybe possible, possible, very likely, and completely possible. The analytical hierarchy process (AHP) is used to determine the weight of the evaluation index [79,80]. The calculation steps of indicator weight are approximately as follows: The farmers' credibility index is determined by questioning the farmers involved in the MVLT in the survey area. Fifteen farmers that are involved in it were invited to form an expert group to solicit their opinions on the role of task allocation, inspection and acceptance, incentive measures, and performance appraisal. The judgment matrix table for the four indicators of the second level in the policy tool according to the scale description (Table 3) was filled in, and 15 judgment matrices were obtained. The judgment matrix table obtained in the above process was inputted into the YAAHP (analytic hierarchy process software V.6.0) 2 . The judgment matrix and weight value of each index were obtained. In order to eliminate the deviation of personal understanding of the index weight, the average value of the index weight in 15 judgment matrices was taken as the final weight.
Set m i as the weight of the index x i . For i, the credibility is C i : where X i is the standard value of the index i. The credibility index intuitively reflects the degree of trust in a policy. The value of credibility is between 0 and 1, which is continuous. The higher the value is, the greater the credibility.

Indicators for the Measurement of Policy Instrument Indices
The following indicators are established for the policy instruments of the MVLT. Each indicator was assigned by the Likert level 5 scale according to the degree of the problem, from minimal to significant.
1. Task assignment: the MVLT has the time node to finish. 2. Inspection and acceptance: the superior established a strict assessment system before, during, and after the event. 3. Incentive measures: according to the completion of the MVLT, there is a certain amount of incentive funds. 4. Performance assessment: the completion of the MVLT shall be included in the annual performance assessment, and those who fail to pass it shall be degraded.
We investigate the lower government's perception of the upper government's policy instruments and ask the former's personnel in the MVLT to complete the questionnaire. The weight of each sub-index of the policy instrument is obtained using a method similar to that for the credibility weight.
Let w i be the weight of the policy instrument indicator p i . For i, the policy instrument value of the indicator is PTI i : The value of the policy instrument index reflects how strongly the upper government pushed the lower government to finish the project, and the value range is [0, 1]. It is also continuous, and a higher value indicates greater efforts from the upper government to promote the implementation of policies.

Administrative Means of Lower Governments and Evaluation of Farmers' Characteristics
The evaluation methods of the division of the administrative means from Table 1 and the peasant household characteristics are shown in Table 4. In the implementation of the MVLT, the situation of each household and the administrative means that they perceived are different. According to the classification of administrative means in Table 1, we requested each household to fill in the specific sheet and subsequently calculate the value of administrative means perceived by each household according to Table 4.
In order to study the impact of individual characteristics of farmers on the credibility of the policy, we used the method in Table 4.
A regression model is constructed by taking the credibility of farmers as the dependent variable and the characteristics of farmers and the administrative means of lower governments as the explanatory variables as follows: Credibility i = β 1 x i1 + β 2 x i2 + β 3 x i3 + . . . + β j x ij + γ j g ij + β 0 + ε i (6) In Formula (6), X ij is the standard value of the characteristic index of the farmer j with the sample i, and g ij is the standard value of the administrative means perceived by the farmer j with the sample i. β 0 is a constant term, and ε i is a residual term.

Classification of Conflict Index
In order to better describe the degree of conflict, the conflict index is divided into five equal parts between 0 and 1, corresponding to the change in conflict from low to high. Fan and Yang [40] classified the credibility according to the credibility scales and intervention (CSI) checklist of Peter Ho [41], which is the basis for the timing and mode of government intervention (Table 5). Following the credibility classification and intervention checklist setting, we obtained the Conflict Degree Intervention (CDI) checklist (Table 6). In Table 6, we divide the degree of conflict into five levels according to the size of the conflict index and propose targeted policy intervention measures for different degrees of the MVLT conflict, providing a basis for timing and intervention measures of conflict governance.

Credibility and Level of Conflict in the Initial Implementation of the MVLT
At the initial stage of the implementation of the MVLT in Q village, the working group of the LID project led by the town government distributed propaganda materials to each farmer; relayed the benefits of the MVLT, the policy of demolition, and the compensation standard; and persuaded them to sign the demolition agreement. At this time, we investigated the credibility of 351 farmers in the project of the MVLT. We also investigated the perception of the policy instruments of 12 staff members in the town government who participated in the MVLT and calculated the value of policy instruments.
We conducted a follow-up survey on the credibility of 351 households participating in the program of MVLT, analyzing changes in the credibility of all households at different stages. Considering household characteristics and the administrative tools used by lower governments, there are three groups. The first group consists of 61 households (Group 1) with high credibility who did not face any administrative tools in the early stages of policy implementation. The second group consists of 242 households who eventually signed the demolition agreement, with 85 households signing the agreement when the government used "soft means" (Group 2-1) and 157 households signing the agreement when the government used "soft means and hard means" (Group 2-2). The third group consists of 48 households with the lowest credibility who did not sign the demolition agreement, regardless of the measures taken by the government. We further analyzed the characteristics of this group and found that 36 households had low income and little social capital (Group 3-1), and their main reason for not relocating was that the compensation amount did not meet their expectations. The remaining 12 households had higher income levels and corresponding social capital (Group 3-2), and their reason for refusing relocation was not the compensation standard but their dissatisfaction with the policy of MVLT.
It can be seen from Table 7 that the credibility of all farmers is 0.4763, and the corresponding risk index is a neutral (upper) level, which is a relatively high level. After ranking the credibility, 20% of farmers (70 households) had the lowest credibility at only 0.2117, which is a medium-low (lower) level, and the corresponding conflict index is 0.7736, which is a medium-high (upper) level of conflict. It demonstrates that the willingness of farmers to participate in the MVLT is very low. Forced participation will inevitably lead to a higher level of conflict. Our survey also shows that the work group is under substantial pressure to implement the MVLT. Note: In order to reflect the change of the credibility level in Table 5 and the conflict index level in Table 6 in a more detailed manner, each level of classification is divided into upper and lower levels.
After measuring the credibility and conflict index in the full sample, it is ascertained that there is an enormous difference in credibility among farmers. This can be divided into three groups, from high to low (Table 7). Among them, 61 households have an average credibility of 0.879 for the MVLT, and the corresponding level of credibility is high (lower). They directly accept the policy requirements and sign the demolition agreement, which we divided into the first group. It is worth noting that the 12 households in group 3-2 have a credibility of 0.0512, they are completely against the MVLT, and the conflict level is very high. According to Formula (2), when the PTI value is fixed, the credibility is inversely proportional to the conflict index. The analysis of the factors affecting credibility will reveal what causes the change in the conflict index.

Individual Characteristics and Credibility of Farmers
The implementation of the MVLT depends on the extensive participation of farmers. Their individual characteristics are different, which determines their varying degrees of willingness to participate and also establishes the credibility of the policy. Individual characteristics are distinguished into whether to live in a policy implementation area, housing structure, income source, income level, and social capital, and these variables have different effects on credibility. The characteristics and credibility of farmers were analyzed via regression. Column (1) in Table 8 shows the relationship between the credibility and the characteristics of the full sample of farmers. The results demonstrate that living in a policy implementation area has a significant negative correlation with credibility, indicating this living situation will reduce the credibility of the policy. In addition, there is a significant correlation between credibility and monthly income; the higher the income, the higher the credibility. There is a negative correlation between the credibility and social capital; the richer the social capital, the lower the credibility. Hypothesis H1 is verified. Robust t-statistics in parentheses; *** p < 0.01, ** p < 0.05.
The full sample is divided into three groups according to the level of credibility in order to find the factors affecting the credibility in more detail. Table 8 (2) is the first group with higher credibility greater than 0.8; Table 8 (3) is the second group with middle credibility between 0.3 and 0.8; Table 8 (4) is the third group with lower credibility than 0.3.
At the initial stage of the policy implementation, the lower government did not undertake the use of any administrative instruments; however, there were 61 households with high credibility of 0.8790, which is at a high (lower) level. Table 8 (2) shows the regression results on the credibility and the individual characteristics of the 61 households, showing the impact of different individual characteristics on the credibility. The most important individual characteristic is whether they live in a policy implementation area, and the p value for this passed the significance test. It indicates that farmers who do not live in a policy implementation area have higher credibility. A further survey of the 61 households supported the regression results, of which 49 were unoccupied, and the owners lived in other cities for many years. Another 12 households moved to the county town, and the cost of demolition can subsidize the cost of a house there. This is why this group has such high credibility in the MVLT.
There are 242 households in the second group. At the beginning of the implementation, their credibility of the policy is not high, ranging from 0.3 to 0.8. Table 8 (3) shows the regression results of the credibility of the second group and the individual characteristics. It can be seen that the credibility of this group is significantly related to the housing structure, income source, and monthly income. This indicates that, for this group, the index of the housing structure, income source, and monthly income will have a positive impact on the credibility.
There are 48 households in the third group. The regression results of the credibility and individual characteristics of this group are shown in Table 8 (4). It illustrates that the credibility of this group is significantly negatively correlated with social capital. The richer the social capital is, the lower the credibility. The credibility of this group is only 0.1742, and the corresponding conflict index is 0.8104. The conflict is at a high (lower) level, and there is a possibility of a conflict developing.

Administrative Instruments and Credibility
At the beginning of implementing the MVLT, the second group's credibility was at a low level. However, when the lower government imposed administrative instruments, the credibility of the 242 households was improved, and this group signed a demolition agreement. Through the follow-up survey on the credibility of the second group and the administrative instruments of the lower government, we found that the credibility of these farmers did, indeed, change. The regression results in Table 9 (1) show that improvements in the credibility are related to all of the three administrative instruments, and there is a significant correlation with hard and soft means. This fully demonstrates that the administrative instruments of the lower government can improve the credibility of these farmers and reduce the risk of conflict. Thus, Hypothesis H2 is verified. Table 9. Regression results of credibility and administrative instruments. Robust t-statistics in parentheses; *** p < 0.01, * p < 0.1.
There were 48 households that did not sign the demolition agreement. The regression results of the credibility of these 48 households and the use of administrative tools are shown in Table 9 (2). It indicates that the credibility of these 48 households is not related to the use of administrative tools by the lower government. At this point, these administrative tools are limited in improving the credibility of this group, indicating that the use of administrative tools is limited. When the lower government cannot improve the credibility of this group or uses all administrative tools to reduce the conflict index, the degree of conflict is highest. Therefore, Hypothesis H3 is verified.
The 48 households can be further divided into two subgroups. The first group consists of 19 households who are "strong nail households" because they strongly oppose the demolition. Among them, 12 households have relatively strong social capital, with close relatives working in government departments or influential public utility departments. This group has a higher income, so the residents' objection to the demolition is not due to compensation but rather their disagreement with the MVLT policy. There are also seven households whose houses were demolished. Because their compensation demands exceed what the government can offer, they refuse to lower their compensation goals. The lower government chooses to use its administrative power to completely clear the houses without their consent, making them the most strongly opposed farmers. The second group consists of 29 households whose income is low and cannot afford to purchase new buildings with the demolition compensation. Therefore, these farmers have been in conflict with the lower government, hoping that the compensation amount can meet their demand for purchasing new buildings in the same area. Otherwise, they refuse to have their property demolished and will do everything to protect their houses. The author calls this group "weak nail households".  (3) and (4)) shows the regression results for the two groups (strong nail households and weak nail households) on policy credibility and changes in administrative tools. Table 9 (3) shows that there is a significant negative correlation between the credibility change in strong nail households and compensation measures. Further investigation found that if the lower government adopts extreme measures to exert pressure, it will inevitably lead to intense conflicts. The author conducted structured interviews with the strong nail group and then assessed their credibility. The credibility value decreased from 0.1742 to almost 0, and the conflict value reached the highest level.
Additionally, Table 9 (4) shows a significant negative correlation between credibility change and soft means for the weak nail households. Their credibility value decreased from 0.2185 to 0.0712, with a corresponding conflict index of 0.9114, which is a high (high) conflict level. This means that the conflict has reached its highest level, and conflict may erupt at any time, endangering social stability. Thus, Hypothesis H3 is verified.

Conflict Pressure after the Policy Stop
The MVLT has been implemented on a pilot basis in different regions of Shandong Province. The dissatisfaction and struggles of affected farmers are made public through various channels. An enhancement in online public opinion began on 12 May 2020, when He Xuefeng published an article titled "Merging villages and Living together, Why should Demolish Farmers' Houses?". On 6 May 2020, the government affairs dynamic column of the Department of Natural Resources of Shandong Province announced that it held an expert seminar and technical regulations on projects such as the Special Plan for the Layout of Villages in Shandong Province. According to the news, the Office of Land and Spatial Planning will "draw up a special plan for the layout of villages in the whole province, guide all localities to complete the layout of villages at the county level, and formulate a plan in the whole province to steadily promoting the for MVLT". It aroused widespread concern 3 and created substantial social pressure on public opinion. Liu Jiayi, secretary of the Shandong Provincial CPC Committee, stated: "Those policies that are being implemented but bring out strong opinions from the masses, those that are being studied and prepared for implementation, and those that have been studied but have not yet been implemented will all be suspended and re-screened." 4 . In this way, the MVLT in Q village was stopped, and the household that had not been demolished was retained and became a nail household.
Since the MVLT was stopped, the conflict index decreased to 0. In order to study the change in credibility after the termination of the MVLT, the authors continue to track and investigate 48 households (including those that have been demolished, strong nail households, and weak nail households). The survey results show that, before stopping the MVLT, the credibility of this group was 0.1750, the conflict pressure was 0.8250, and the level of conflict was high. With the disappearance of the policy tool, the credibility is adjusted to 0.6943; it does not disappear but it also does not increase close to 1. The conflict pressure is 0.3057, which is at the neutral (lower) level. This shows that, even if the MVLT is abolished, the negative impact of the policy on farmers will not disappear completely, and there is a lag effect of administrative instruments that buries hidden dangers for the government's follow-up policy formulation.

Discussion
Based on the credibility thesis and underlying theory, this paper aimed to explain the key to conflict generation and resolution. The authors also examined the willingness of farmers to participate in the MVLT quantitatively by calculating the credibility that reflects the degree of the conflict index. In previous studies, it was assumed that farmers had complete freedom of choice, and their individual characteristics and preferences determine the credibility and, thus, establish their own behavior choices. The authors ascertained that those farmers who moved out of the village received compensation to improve their income, and they had the highest credibility. For those farmers who live in villages, the reasons for the different credibility are due to the different preferences and livelihood levels. It is necessary to implement refined and different policies.
In the implementation of public policy, the credibility of farmers is not only endogenous but also affected by the external environment. In the implementation of the MVLT with low credibility, the lower government imposed different administrative instruments to increase the credibility of 242 households that finally signed the demolition agreement, accounting for 68.9% of the total number. That means that the lower government imposed different administrative instruments or combinations of those with a comprehensive investigation of the characteristics of different farmers, which will increase farmer credibility and achieve the goal of conflict resolution. Based on the study of the MVLT's implementation process, this paper calculates the change in the credibility of the farmers involved and finds that the administrative instruments of the lower government changed the credibility of some farmers. It also effectively depicts the change in the strength of governance means and accurately describes how the conflict of the MVLT is resolved using the administrative instruments of lower governments. This will bring new enlightening significance to the study of conflict theory. For those farmers who do not have social capital, the government's forced demolition not only fails to improve their credibility but also brings the possibility of conflict at any time, endangering social stability. Those farmers whose family income is not high and cannot afford new buildings with no social capital will fight for their lives in the face of forced demolition. In this situation, the lower government cannot use administrative instruments to achieve policy objectives indefinitely under the unchanged policy tools. The role of upper-government policy tools is also very important as it is the key to promoting policy implementation and resolving conflicts. The nail households after the termination of the policy are precisely the manifestations of conflict resolution.
Theoretically, the authors believe that an important contribution of this paper is to introduce the "credibility thesis" into the analytical framework of policy implementation and to prove that the conflict is the result of the interaction between the credibility of farmers and the upper government's policy tools. Additionally, the administrative instruments of the lower government affect the credibility of farmers and become an important measure for resolving the conflict. The authors also find that under the same land expropriation policy, the difference in farmers' preferences and livelihood status leads to a difference in credibility, which results in different behaviors in the implementation of the MVLT. For the methodology, the authors developed the method of how to measure conflict pressure, policy tools, administrative instruments, and conflict index as well as evaluate the relationship between credibility, government behavior, and conflict in a quantitative way. This provides theoretical and methodological support for the emergence and resolution of public policy conflicts and also contributes by establishing a basis for adopting differentiated policy interventions.

Conclusions
This paper began with an analysis of the credibility and willingness of farmers, the upper government, and the lower government to implement the policy of MVLT. It also established an explanatory framework for policy formulation, implementation, and conflict generation; defined the connotation of farmers' credibility, policy tools, and administrative instruments; and constructed a corresponding index system and quantitative measurement methods. A unified and accurate description of conflict generation, conflict intensity, and conflict resolution was provided. Through the case analysis, the following key conclusions can be drawn: Table 7 shows that there is a significant variation in the trust levels of farmers towards the MVLT, with credibility levels ranging from the lowest to the highest. The author found that the farmers who had already relocated had the highest credibility levels. For those living in the village, the variation in credibility levels was due to differences in farmers' preferences and livelihoods ( Table 8).
The use of administrative tools by lower governments can change the credibility of farmers and effectively resolve conflicts. The degree of conflict caused by the MVLT is neutral (upper); however, the 20% of farmers with the lowest credibility levels contribute more to the emergence of conflicts, and the conflict index is at a high (upper) level (Table 7). Based on a comprehensive investigation of the characteristics of different farmers, lower governments choose different or combined administrative tools (such as "hard means" or "soft means") to improve their perception of profit and loss, conflict, and institutional change, thereby increasing their credibility levels ( Figure 3, Table 9) and achieving conflict resolution.
The ability of lower governments to resolve conflicts using administrative tools is limited. When all administrative tools have been used but the credibility levels of farmers cannot be increased or the conflict index cannot be reduced, the degree of conflict reaches its highest level.

Data Availability Statement:
The datasets analyzed during the current study are not publicly available due to the personal private information involved but are available from the corresponding author on reasonable request.