Does Supervision Down to the Countryside Level Benefit Rural Public Goods Supply? Evidence on the Extent of Households’ Satisfaction with Public Goods from 2005 to 2019

Abstract

This paper mainly studies the effect of village-level power supervision reform on the rural supply of public goods. In this paper, the panel data of 100 villages in five provinces of China from 2005 to 2019 are used to identify and analyze the impact of the supervision policy on the supply of rural public goods. The study adopted PSM-DID as the identification strategy to alleviate the endogenous problem of the model; it found that the supervision policy significantly improved the supply of public goods in villages, and increased the satisfaction of villagers with the village economic base and public social services. There is regional heterogeneity in the improvement effect of policy on public goods supply, which mainly reflects the regions with weak economic development. This policy mainly promotes the optimization of public goods supply by strengthening villagers’ public participation, and promoting the improvement of the current situation of public goods by restraining village cadres’ duty behaviors. It is also found that the coupling of clan power and village formal authority can inhibit the positive effects brought by the policy, and only in areas with strong clan power can the negative effects be mitigated. Moreover, there is an alternative relationship between the assessment pressure of village cadres and the supervision force sent to the countryside, which will form a situation of excessive supervision and inhibit the optimization of village public goods. The conclusion of this paper provides empirical support for the view that “top-down external institutional supervision and bottom-up internal democratic supervision should be effectively integrated” in the theory of village power supervision.

1. Introduction

The self-governance of rural villagers, with the election of a village committee as the core content, is regarded as a “democratic experiment” carried out from bottom to top in China. This democratic practice provides a new research direction for the governance of rural society in China. Grassroots democracy is an important component of socialist democratic politics with Chinese characteristics, and the grassroots democracy system in our country is pioneered and led by “villagers’ self-government”. Characterized by “democratic election, democratic decision-making, democratic management, and democratic supervision”, “villager autonomy” has greatly enhanced the Chinese people’s consciousness of democracy and rights, promoted democratic participation, and enhanced trust in the system [1,2,3]. It also plays an important role in promoting the supply of public goods in rural China and ensuring the modernization of rural society [4,5]. However, with the rapid economic and social development seen in recent years and the great strides made in the pace of urbanization, the operational effectiveness of “villager autonomy” has not been continuously enhanced along with the development of the rural economy and the improvement of farmers’ lives, but has declined, or even been paralyzed. The conflicts between grassroots cadres and groups in rural society are still prominent and cannot be solved only through the institutional framework of “villager autonomy” [6].

At present, “villagers’ self-governance” established by the Constitution and the Village Group Law on the basis of administrative villages has largely lapsed. With the tightening of financial budgets after tax reforms, village cadres must redeploy limited financial resources, while the administrative cost required in the management of the daily affairs of villages is relatively fixed, and unlikely to be further reduced, so tax reform often weakens the positive effect of villagers’ self-governance on administrative efficiency [7]. In order to realize the linkage of supervision between inside and outside the village, rural society began to try to establish an independent professional supervision organizational system for village affairs. The discipline inspection department combined with local reality made full use of the internal resources of the village, and established a village affairs supervision committee by selecting election personnel from the old petitioners or those with high moral values in the village, thus realizing a power balance among the organizations.

Due to the complexity of rural grassroots affairs and the limitation of the authority of the village affairs supervision committee, the systematic requirements and “fragmented” reality of the rural grassroots supervision system, the effectiveness requirements of rural grassroots supervision and the insufficient supervisory power of the village affairs supervision committee, along with other important contradictions, the effectiveness of rural grassroots supervision is reduced, and the independence of the village affairs supervision institutions is insufficient [8,9]. The problem of the “bottom-up” supervision system is that it relies on the supplement of “top-down” supervision, which lays the foundation for the reform of village-level power supervision—supervision to the countryside.

Many studies have investigated the influence of power supervision on public governance. On the one hand, power supervision can restrain the behaviors of those in power through system construction, and enhance the effectiveness of supervision through economic and cultural measures, so as to reduce corruption [10,11]. Strengthening cooperation and docking between grassroots governments and non-governmental organizations (informal organizations) can enhance public participation and thus achieve better public environmental and health governance [11,12,13]. On the other hand, the characteristics of top-down supervisors are closely related to the effectiveness of supervision. Chinese scholars have put forward a new point of view. With the “regime’s coming to the countryside”, it has brought more public resources to villages. Coupled with the constraints of democratic supervision, its internal role of helping the supply of public goods is reflected. The villages and towns establish “a set of comprehensive management methods with the index system as the core, the responsibility system as the basis and the assessment system as the power, radiating the formation of target management network, in order to obtain the best administrative efficiency”. This increases the supervision over the members of village-level organizations, and enables them to partially undertake the function of providing public goods required for social security entrusted by the government [14].

However, imbalances in power supervision will lead to corruption, thus reducing the voluntary supply and trust of the public [15,16]. After state power enters the village level in various forms, it will promote the administration of village cadres and the semi-administration of village governance, which can also explain the internalization of village funds, including the solidification of the fund expenditure structure, the rise of village governance costs, and the squeezing of administrative affairs into village daily affairs [17].

Although national laws and regulations attach great importance to the reform of village-level power supervision and have achieved remarkable results, many studies have shown that the effectiveness of supervision in China’s rural communities is not high. The available literature attributes China’s bias to two aspects. First, the independence of supervisory bodies is insufficient. The complexity of rural grassroots affairs and the limited authority of village affairs supervision committees, the systematic requirements and “fragmented” reality of rural grassroots supervision system, the effectiveness requirements of rural grassroots supervision and the insufficient supervision power of village affairs supervision committees, along with other important contradictions, reduce the effectiveness of rural grassroots supervision, and the independence of village affairs supervision institutions is insufficient [8,9]. Second, hitchhiking behavior is difficult to solve. Supervision to the countryside strengthens the administrative characteristics of village cadres (which is the main body of village governance), weakens the autonomy characteristics, and induces the generation of grassroots formalism [18]. The ability of village cadres to organize farmers is reduced, which makes farmers unable to take collective actions. In the face of national resources being sent to the countryside, farmers are more likely to display “free ride” behavior, which makes the corresponding policy of benefiting agriculture hard to implement, and this has a great negative impact on the supply of public goods in villages [19]. From both perspectives, rural governance, including formal democratic governance and informal customs, may play an important role in shaping the security and integrity of farmers’ land rights.

However, in many developing countries, the effectiveness of rural governance systems has not been fully realized due to the weakness of local governments, the distorted interpretation or incorrect implementation of legal documents, and other factors [20]. In this context, a series of different policies and measures have been proposed, including attempts to improve the allocation of rural public resources by improving local resource management models, government institutions and infrastructure [21,22,23]. China has also never stopped exploring ways to improve the village governance system.

During the period of supervision, a vertical management system of county and township committees was established. The village affairs supervision committee enables good cooperation and connection with the superior government and the discipline inspection commission, embeds the external supervision force into the village supervisory committee, and promotes the independent situation of the village supervisory committee, the village party organization and the village committee in the rural society. Specific practices include: First, the township discipline inspection commission and the director of the supervision office appoint the village discipline inspection supervisor (the village discipline inspection commission), and one of them holds three positions, namely, “the village discipline inspection commission, the village supervision committee director, the discipline inspection supervisor (called the clean government supervisor in some areas)”. They are directly responsible to the township discipline inspection commission. Secondly, raise the salary of members of the supervisory committee. In order to solve the problem of the poor treatment of discipline inspection commissioners, towns and townships have raised their own funds to subsidize each discipline inspection commissioner on the basis of monthly subsidies from the county’s finances. Third, the village begins to standardize the collective public affairs process. To realize the standardization of the management process of village social services and public affairs, it is necessary to carry out the process strictly in accordance with the project proposal, with project consultation, project decision-making, reviewing and recording, organization and implementation, and project supervision, and to supervise the whole process of the village affairs decision before, during and after the implementation. The specific content will be detailed in the following article. This reform addresses a wide range, and has had great influence. In the past five years, discipline inspection and supervision bodies have dealt with 48,000 cases of corruption and work style issues in the field of rural revitalization, gave 46,000 Party disciplinary and administrative punishments, and 882,000 villages and communities were inspected by the central government for reform (https://www.ccdi.gov.cn/toutiaon/202210/t20221027_227265.html, accessed on 20 March 2023).

The purpose of this paper is to investigate the effects of supervision on farmers’ satisfaction with the supply of public goods, and to determine the effects of villager participation, clan power, and village cadre assessment on regulating the effectiveness of supervision. In this study, the data of five large-scale household surveys in Jiangsu, Sichuan, Shaanxi, Hebei and Jilin provinces from 2005 to 2019 were selected for the empirical analysis, and the identification strategy combined with DID and propensity score matching was used to alleviate the endogenous problem of the empirical model, whereby the interaction was estimated.

The relevant findings and implications provide new insights relevant to promoting the reform of the rural or community governance system and the integration of rural public goods supply systems in China and other developing countries. This paper attempts to contribute to this literature in three ways.

First, our research attempts to address the theory of power supervision, in the field of public governance, to strengthen the effect of supervision on the supply of rural public goods via an in-depth study. The research shows that the coupling of top-down supervision and village internal supervision can improve the willingness of villagers to participate in public affairs independently, and regulate the duty behavior of village cadres, so as to improve the efficiency of public goods supply and villagers’ satisfaction, which is crucial for rural revitalization. Second, we supplement this with research content. Although a few studies have considered power supervision when studying the factors affecting public governance [24,25,26], and some scholars have proposed that a dual leadership system with “cover” can alleviate the distortion of incentives brought about by the subjective supervision mechanism through studying the assessment and supervision of China’s township governments, these studies have not paid attention to the particularity of acquaintance societies in rural areas. “Cover-up” supervision and assessment is not applicable in rural society, as it ignores the auxiliary role played by informal organizations in governance [27]. However, this paper discusses the regulating role of clan (non-governmental organizations) and village cadre assessment (formal system supervision) in the supply of rural public goods, which has rarely been considered in previous studies. This study improves the internal adjustment mechanism of the influence of power supervision reform on public governance, and puts forward the idea of and empirical evidence for the substitution relationship between external supervision force and internal institutional constraints of villages, finding that villages should relax their original internal institutional constraints while strengthening the external supervision force. Three, we improve the research methods. The relevant existing literature lack long-term panel data and rigorous empirical analyses. In this paper, the panel data of 100 villages in five provinces of China from 2005 to 2019 are used to investigate the impact of strengthening vertical power supervision at the village level on the supply of rural public goods by applying the supervision policy to the countryside as a quasi-natural experiment. The sample size is large, and the differential difference method is used to overcome the estimation bias found in previous studies, as well as enhance the reliability of regression results.

The rest of this paper is arranged as follows: The second part is the policy background and theoretical analysis. The third part introduces the research design. The fourth part gives the estimation results and analysis, including baseline regression results, heterogeneity analysis and robustness test. The fifth part examines the functional mechanism and adjustment mechanism of the policy set out to improve the supply of rural public goods; the last part contains the conclusion and policy implications.

2. Policy Background and Theoretical Analysis

2.1. Policy Background

Villager autonomy is the internal mechanism of village public power supervision. Under this system, the supervision mode of village public power is shown in Figure 1. Village Party organizations, under the leadership of grassroots government Party organizations and composed of village Party members, are the core element of leadership in rural grassroots organizations. They coordinate the relationship between various rural organizations, coordinate and deal with the interests and contradictions of various elements, and ensure the steady and rapid development of the rural economy and social harmony and stability. The villagers’ committee is an autonomous organization elected by the villagers to handle the public affairs and public welfare undertakings of the village, mediate civil disputes, and assist in maintaining social order. The village affairs supervision committee is a village affairs supervision body established according to law. It is responsible for the villagers’ democratic financial management, and the implementation of the system of making village affairs public. Its members are elected by villagers’ meetings or representatives’ meetings from among the villagers.

This supervision mode introduces village public power into the social internal supervision system. Although under the leadership of the Party organization at the village level, the village committee is bound by the two supervision paths of intra-Party supervision and democratic supervision, and the village cadres still have the opportunity to get in touch with superior cadres. Under the effects of the unified village power structure, the administrative target management responsibility system leads to the isolation of, and even contradictions between, the multiple supervision systems of village cadres, to a certain extent. Giving village cadres the opportunity to collude with higher-level government officials results in the loss of villagers’ welfare [28].

After the tax reform, a large amount of funds and resources poured into rural society, and the phenomenon of village cadres using their power for personal gain emerged to a great extent, with some receiving kickbacks in land acquisition or demolition, and lining their own pockets via the village’s “three capital” management process, which led to conflicts between cadres and groups as well as mass incidents, prompting the state to take measures to strengthen supervision over the power of village cadres.

As a reform of the village-level power supervision system, supervision of the countryside is an important means for the higher government to adjust and assist the villagers’ self-governance, which is of great significance in maintaining rural social stability and ensuring the supply of public services. The implementation of the reform intended to supervise rural residents is divided into three stages (Table 1):

(1)

2004–2007—The establishment of a village-level internal supervision organization under villagers’ self-governance.

The earliest supervisors sent to the countryside tried to supervise the power at the village level by mobilizing the internal forces of the village. Here, the villagers themselves elect highly respected members of the village to be the directors and members of the village supervisory committee, who listen to public opinion and supervise the daily behavior of the village cadres. However, this kind of supervision mainly relies on the internal strength of the village, and has not shown much effectiveness. The reason is that the village affairs supervision committee is also organized and managed by the two village committees and lacks independence;

(2)

2008–2012—Village discipline inspection commissioners were set up to send villagers to the countryside on behalf of superior government supervision forces.

In Huaihua City, Hunan Province, in response to the increasingly serious problem of village cadre corruption and to strengthen the construction of honset and grassroots Party conduct in rural areas, the Mayang Miao Autonomous County discipline Commission set up discipline inspection commissioners at the village level, and appointed Party conduct supervisors at the township level. By using the deputy secretary of the village branch as a part-time supervisor, they were able to save the expense of hiring full-time supervisors [25];

(3)

From 2012 to now—A discipline inspection mechanism of “town discipline inspection commission + village supervision committee” has been established to realize the coupling of external supervision and internal supervision.

Jiangsu, Sichuan and other local governments issued official documents emphasizing that “the annual assessment of the village discipline inspection commissioners shall be carried out by the town Party committee and the town Discipline Inspection Commission, and the discipline inspection commissioners shall report on the situation of hard work and clean government and accept quantitative assessment”, which is equivalent to the reduction of power supervision over the township’s government to the village level, whereby the village discipline inspection commissioners become the representatives of the township’s government at the village level (https://www.shehong.gov.cn/web/shehong/xzdt/-/articles/9867823.shtml, accessed on 10 May 2022. https://www.shehong.gov.cn/web/shehong/xzdt/-/articles/9865378.shtml, accessed on 7 May 2022. http://www.dafeng.gov.cn/art/2015/5/4/art_24999_3572003.html, accessed on 12 May 2022). It can be seen that the timescales and practices of village-level power supervision reform in different regions are different.

Table 1. Supervision of the reform and changes in rural areas.

Time	Model	Innovation Characteristics
2004–2007	“Internal selection of supervisors + government assessment”	Select members from the petitioning villagers; village discipline inspection committee members concurrently serve as the directors of the village supervision committee, and the annual assessment is directly supervised by the town Party committee and the discipline inspection committee
2008–2012	“Posted supervisor”	Late initiation of reform; dispatched “village discipline inspectors” to administrative villages to supervise on behalf of higher discipline commissions
2012–now	“Internal selection of supervisors + government assessment”	The members of the supervision committee shall be selected from the village elites and verified by the township Party committee and government; a discipline inspection mechanism of “town discipline inspection commission + village supervision committee”

Table 1. Supervision of the reform and changes in rural areas.

Time	Model	Innovation Characteristics
2004–2007	“Internal selection of supervisors + government assessment”	Select members from the petitioning villagers; village discipline inspection committee members concurrently serve as the directors of the village supervision committee, and the annual assessment is directly supervised by the town Party committee and the discipline inspection committee
2008–2012	“Posted supervisor”	Late initiation of reform; dispatched “village discipline inspectors” to administrative villages to supervise on behalf of higher discipline commissions
2012–now	“Internal selection of supervisors + government assessment”	The members of the supervision committee shall be selected from the village elites and verified by the township Party committee and government; a discipline inspection mechanism of “town discipline inspection commission + village supervision committee”

After the initial tests, the reform was gradually rolled out nationwide. The Opinions of the CPC Central Committee and The State Council on Focusing on Key Work in the Fields of Agriculture, Rural Areas and Farmers to Ensure the Realization of a Moderately Prosperous Society in an All-round Way on Time, issued in early 2020, pointed out that the communication and cooperation between grassroots discipline inspection and supervision organizations and village affairs supervision committees should be strengthened, and effective connections should be established to form a joint supervision force.

Supervision over the countryside is essentially a supplement to the villagers’ self-governance system, and a reform of policy set out by the grassroots government to change the prominent problems in villagers’ self-governance. This policy does not deny the rationality of villagers’ self-governance, but tries to solve the existing problems therein.

Village-level power supervision has the following characteristics: First, the focus of village public power falls on village collective affairs, and the strength of both internal and external supervision increases. The purpose is to give full play to the synergistic effects of internal and external supervision, and realize the coupling of external intra-party supervision from the top down with self-supervision of the masses from the bottom up. Second, village rights supervision has gradually changed from the mode of internal and external joint supervision to the mode of internal and external dynamic supervision. The intervention of external forces in rural society is a supportive institutional change, and its realization requires the benign cooperation and interaction of multiple subjects, such as the internal organizations of rural society, the two village committees and the villagers, so as to promote the optimization of village governance in cooperation and conflict. In this process, we must clarify the functions and responsibilities of internal and external forces, and establish a corresponding system to ensure the coordination of internal and external forces.

2.2. Framework

American scholar Ostrom’s multicentric governance theory takes the supply of public goods as the core of governance, aims to improve the supply efficiency of public goods, and emphasizes the pluralism of governance subjects. This theory holds that the government is not the only provider of public goods, but is one of the main providers. It advocates for the combination of government, market and non-governmental organizations. The more widely accepted view in the academic circle is the “multiple supply” theory, incorporating government supply, private sector supply, autonomous sector supply, and the individual supply of farmers. This paper focuses on the “last kilometer” of the supply of rural public goods, which is the key link in determining the effectiveness of the supply of rural public goods. The main body of the supply system is village-level organizations and farmers, who need to complete various projects and establish a proper communication and coordination mechanism.

At present, the general framework for studying the influence of institutions on governance is “system–behavior–performance”, that is, the SBP analysis framework. Different public goods suppliers will show different responses to external policy impacts, and these responses will also affect the final effect of public goods supply. Therefore, the following analysis framework is established in this paper, shown in Figure 1.

Supervision to the countryside must be procedural, standardized and institutionalized at the county and township levels, in order to build an inspection, supervision and accountability system. Power supervision must be incorporated into the village organization, that is, the township government, in order to strengthen the vertical power supervision of the village organization.

In traditional rural society with low mobility and a lack of formal systems, village power supervision is key to realizing the adjustment of rural relations. Village democratic supervision and the villager self-governance system determine the boundaries of state power intervention, but in the context of certain governance aims and policy choices, township governments will also use administrative power to control village rights, and assume the responsibilities of guiding, managing and supervising village cadres (village-level rights subjects). For example, village democratic governance has been incorporated into the institutional reform of supervision by the special office of the discipline inspection commission. This is conducive to the formation of a benign and interactive mode of village administration supervision [29]. As a kind of strategy choice for township governments in their specific environment, the village-level administration also represents an effort to overcome the “last kilometer” dilemma of rural governance [30]. However, with the strengthening of the formal supervision system, the corresponding functions of this supervision system also began to differentiate, which may restrict the village cadres from exercising their responsibilities and powers, allowing them to only focus on dealing with the investigations and tasks of the superiors, or even preventing them from exercising their rights [31,32]. The subjectivity of villagers and their enthusiasm in participating in the governaanzheci chunshu nce of village public affairs has declined [19]. Therefore, this paper focuses on the impact of this policy on the supply behavior of village cadres and villagers.

Domestic and foreign studies have also found that other formal institutions and informal institutions in the village related to the supply of public goods play a role in regulation. This article includes village cadre and clan power assessments. Strengthening the construction of grassroots cadres in rural areas is the key to solving the problems related to agriculture, rural areas and farmers, which is the top priority of the Party. As the personification of the government, the performance level of rural grassroots cadres determines the effectiveness of the government’s management of national economic, social and political affairs. As an influential informal system, clans play an important role in promoting rural labor flow, alleviating the income gap within villages, alleviating rural poverty, and promoting the development of rural enterprises [33,34,35,36]. Following the “functionalism” analysis path set out by Freedman (1965) in his classic study on clans in South China, some studies have come to essentially regard the realization of rural public goods supply as the primary reason for the existence and function of clans [37,38]. Some studies have revealed this interaction, specifically examining how clans constrain formal authority and thus promote the supply of rural public goods [39], but some believe that this influence is negative [40], and some even suggest that there is no obvious correlation between them [41] (Figure 2).

2.3. Theoretical Analysis

In order to explore the possible influence of supervision on the supply effect of village public goods, this paper analyzes from four points of view: the participation of village cadres in supply, the participation of villagers, the power of clans and the assessment of village cadres.

• Supervision to the countryside will have a restraining effect on village cadres, standardize the duty behavior of village cadres, and affect the personnel base of the supply of rural public goods. The reform of taxes and fees has changed the operational mode of grassroots political power, and the state has transferred public service resources downward through the “project system”. In the “county-based” project supply mode, the village and village levels do not obtain corresponding authorization due to the downward movement of funds, so after the reform of taxes and fees, the governance power of village cadres is weakened, and they lack the motivation to participate in project dispute mediation [30]. The restraint and supervision effects on the system brought about by supervision to the countryside can effectively resolve the “negative” work attitudes of village cadres. For example, supervision over the countryside standardizes the behavior of village cadres and reduces the occurrence of “small and micro” corruption incidents at the village level. Particularly in the period of project systems, towns and villages often sign letters of responsibility with village cadres. This comprehensive management method will increase the initiative of village cadres, thus reducing the imbalance in project supply and promoting the supply of public goods [42,43];
• The policy of sending supervision to the countryside guarantees the villagers’ rights to democratic supervision, increases villagers’ participation, and affects the mass base of rural public goods supply. In areas where the government is absent and national resources are seldom sent to the countryside, villagers’ voluntary supply formed by the clan network is the main force affecting the supply of rural public goods [44]. Some studies directly assess the provision of rural public goods as the main reason for the existence and function of village clan groups from the perspective of functionalism [38]. However, existing studies have found that the impact of formal institutions or new policies can play a role through the trust compensation mechanism, which can not only significantly increase farmers’ willingness to supply public goods [45], but even improve the supply level of actual public services [46]. After supervision is sent to the countryside, the village discipline inspection committee concurrently serves as the director of the village supervision committee, which expands the channels for villagers to offer suggestions, and guarantees the power of democratic supervision within the village. This practice can enhance villagers’ trust in the top levels of government, as well as villager organizations, and promote their participation in the village’s public affairs and public goods supply projects [47];
• As the representative of the informal rural system, the clan will constrain formal authority and play a regulating role in altering the effects on the supply of public goods brought about by supervision over the countryside.

Clan groups can more easily collect villagers’ opinions, consider villagers’ demands for public goods, apply for resources, and effectively integrate the resources required for village public goods supply projects. The collective punishment mechanism can help to effectively overcome the moral problems of village cadres, keep the supply of public goods from deviating from villagers’ wishes, and promote the implementation of public goods supply projects [44,48,49]. Moreover, village cadres are associated with clan names, and this can help persuade clan members to increase their investment in village public goods [34]. However, with the increase in social mobility, a large number of farmers go out to work, stay away from the countryside, and even settle down in the city all year round, and the function of social security and public organization in the clan network thus begins to weaken, making its role in supervising and motivating village cadres becomes almost insignificant, resulting in a decrease in the supply of public goods and a mismatch between supply and demand [48]. This inhibits the internal improvement effect of clan forces on the supply of public goods at the village level. After the supervision is imposed, if the village cadre holds a clan surname, they will not only be subjected to the restrictions of the superior government, but will also be subjected to the restriction of the internal forces within the village, that is, the accountability brought about by the clan surname. However, the accountability imposed on such clans with big family names may not be optimal, and will be affected by the decline in clan power and dysfunctional clan security. As a result, the public goods provided by village cadres will not meet the actual needs of villagers, which will inhibit the improvement of the actual supply of village public goods.

4.

The assessment of village cadres, as another manifestation of formal system constraints, will increase the duty pressure of village cadres and strengthen the formalism of grassroots governance, thus affecting the supply of rural public goods.

Top-down administrative examination not only enhances the pressure associated with tasks, but also leads to diversified methods of reducing the burden at the grassroots level. Rural grassroots governance under systemic pressure will display “egoism”, that is, in order to complete the important tasks “apportioned” by superiors consistently and reduce the risk of being held accountable, the grassroots cadres will “selectively execute” the tasks of the superiors according to their own interests, and this may even lead to the grassroots officials only undertaking superficial work and engaging in formalism to meet the requirements of the superiors [50,51]. The incentive formed by the supervision policy can easily have a negative effect on village governance via the assessment of village cadres. Compared with economic indicators, the appraisability of public service indicators is relatively weak; because the appraisability is relatively weak, there are strong incentives to do superficial work to meet the assessment requirements of superiors, or to partake in formalism, and even fraud. Supervision over the countryside is a kind of strong incentive for work and agriculture. The state provides farmers with basic public goods through large-scale transfer payments to the countryside. A large number of grassroots governance operations with poor appraisability can be included in the hard assessment, showing a combination of strong incentive and weak appraisal, leading to the use of many of the rural government’s resources in ineffective governance matters, and the rapid increase in grassroots governance disorder. As a result, problems such as reductions in the supply of public goods and mismatches between supply and demand appear, and the effectiveness of public goods supply is weakened.

Based on the above analysis, the research hypotheses of this paper are proposed, as follows:

H1. 

The supervision of the rural policy has a promoting effect on improving the supply of rural public goods.

H2. 

Supervision of the countryside mainly promotes the improvement of the supply of village public goods by restricting the duty behaviors of village cadres and increasing the public participation of villagers.

H3. 

The coupling of clan power and formal authority will inhibit the effect of the supervision policy on improving the supply of rural public goods.

H4. 

The hard assessment of villages will inhibit the strong incentivizing effect of the supervision policy, leading to a negative change in the effectiveness of the supply of rural public goods.

3. Research Design and Methods

This paper explores the effect of the supervision policy on the supply of rural public goods, that is, the changes in farmers’ satisfaction with different rural public service projects after the implementation of the supervision policy.

3.1. Empirical Model Design

In order to test the above hypothesis, this paper establishes the following benchmark model:

$${\mathrm{Y}}_{{}_{ijt}}={\alpha }_0+{\alpha }_{1}di{d}_{it}+X_{it}^{\prime }{\beta }_1+Z_{ijt}^{\prime }{\beta }_2+{\lambda }_i+{\gamma }_t+{\epsilon }_{ijt}$$

(1)

Model (1) is a bidirectional fixed effects model with a difference-in-difference recognition strategy. DID can alleviate the endogeneity of the model, while the bidirectional fixed effect of time–place can control the effects of other missing variables that change with time and place on the estimation coefficient, increasing the credibility of the estimation results.

Among these, ${\mathrm{Y}}_{ijt}$ is the explained variable and represents the effect of the supply of rural public goods. The “Farmers’ satisfaction with the supply of different types of public goods” entity in the questionnaire is used as the proxy index. This paper focuses on the analysis of four types of rural public goods, namely, education, medical treatment, irrigation and drainage facilities. $di{d}_{it}$ represents the independent variable; it refers to the interaction item of policy implementation, whereby $di{d}_{it}$=$trea{t}_i\times pos{t}_t$. Here, $trea{t}_i$ refers to whether village i is already under supervisor—it equals 1 if the answer is yes, and otherwise it is 0; $pos{t}_t$ is the time dummy variable, which equals 1 if the policy has already come into effect, and otherwise it is 0.$X_{it}$ is a set of village-level control variables affecting the efficiency of rural public goods supply, including village total population, the amount of cultivated land, village per capita net income, village total income, village cadres’ education level, work and business experience, the holding of a big surname or not, and other basic characteristics of village economy and social dexelopment. $Z_{ijt}$ represents the set of peasant household-level control variables that affect the efficiency of rural public goods supply, such as Gender, household registration, age, education level, being a party member or not, being in a village cadre or not, the amount of farmland in the family, etc. ${\lambda }_i$ represents the village fixed effect that is used in controlling individual factors that do not change with time; ${\gamma }_t$ represents the year fixed effect, used to control time factors that do not vary with individuals; ${\epsilon }_{ijt}$ is the random error. ${\alpha }_1$ is the focus of this paper, reflecting the effects on public goods supply brought about by the policy of supervision over the countryside (SDC reform).

Since the multivariate linear model estimated by the ordinary least square method (OLS) struggles to address the possible sample selection bias and the endogeneity of missing key variables, this study mainly adopts the propensity score matching method (PSM) to carry out the robust causal inference of the empirical relationship between supervised rural delivery and the effects on rural public goods supply.

In order to further explore the mechanism of the effect of the supervision policy on rural public goods supply, this paper uses the method of Baron and Kenny for testing [52], and sets out the following model:

$$Channe{l}_{ijt}={\alpha }_3+w_{1}di{d}_{it}+X_{it}^{\prime }{\beta }_3+Z_{ijt}{\beta }_4+{\lambda }_i+{\gamma }_t+{\epsilon }_{ijt}$$

(2)

$$Y_{ijt}={\alpha }_4+w_{2}Channe{l}_{ijt}+X_{it}^{\prime }{\beta }_5+Z_{ijt}{\beta }_6+{\delta }_i+{\gamma }_t+{\epsilon }_{ijt}$$

(3)

Model (2) and Model (3) are also bidirectional fixed effects models. Model (2) represents a test for the existence of a conduction mechanism. $Channe{l}_{ijt}$ represents the conduction mechanism variable, including the function behavior of village cadres and the participation of villagers. The remaining variables are consistent with Model (1). If the coefficients of $di{d}_{it}$—$w_1$ and $w_2$ are significant, it shows that the proposed transmission mechanism does exist.

In order to further explore the moderating effects in the assessment of village cadres and the influence of a big clan name on the policy of supervision in the countryside, this paper makes adjustments on the basis of the main effect model (1), and builds the following econometric model:

$$Y_{ijt}={\alpha }_0+{\alpha }_{1}di{d}_{it}+T_{it}\theta +X_{it}^{\prime }{\beta }_1+Z_{ijt}^{\prime }{\beta }_2+{\lambda }_i+{\gamma }_t+{\epsilon }_{ijt}$$

(4)

$$Y_{{}_{ijt}}={\alpha }_0+{\alpha }_{1}di{d}_{it}+T_{it}\theta +\omega di{d}_{it}·T_{it}+X_{it}^{\prime }{\beta }_1+Z_{ijt}^{\prime }{\beta }_2+{\lambda }_i+{\gamma }_t+{\epsilon }_{ijt}$$

(5)

The paper uses a new addition, $T_{it}$, as the moderating variable in model (4), including the most important assessment index of village cadres and whether the village cadres contain a big-surname family, which are used to measure the constraints of the assessment of village cadres by the overarching government (formal system) and the constraints of the supervision of big-surname families (informal system). On the basis of model (4), model (5) is supplemented with the interaction term of the explanatory variable and the regulating variable.

If the coefficient of the main effect (${\alpha }_1$) is positive, the coefficient of the interaction (w) will also be positive; it can thus be stated that the moderating variable has a significant positive effect on the influence of supervision within the countryside and the satisfaction with public goods supply. If the interaction (w) coefficient is negative, it indicates that the moderating variable weakens or inhibits the influence relationship between the supervision of rural delivery and the degree of satisfaction with public goods supply. In other words, the supervision of the rural delivery policy and the moderating variable show an obvious substitutive relationship in affecting the degree of satisfaction with public goods supply. Otherwise, the conclusion is reversed.

3.2. Variable Selection

(1)

Dependent Variable

With reference to common practice in the existing literature [10], this paper assessed the satisfaction of farmers with four types of public goods, namely, rural education, medical treatment, irrigation facilities and drainage facilities. Among these, rural education and medical care represent social security public goods, while irrigation and drainage facilities represent rural economic infrastructure public goods. All observations were assigned scores as follows: 1 = satisfied (including very satisfied and relatively satisfied); 0 = not very satisfied (including very dissatisfied and not satisfied).

(2)

Independent Variable

A dummy variable of DID was constructed to measure the net effects of policies in the multi-period differential model, which means that supervised visits to the countryside occurred in the village where the samples were located. “Whether the village discipline inspection committee concurrently serves as the director of the village supervision committee and is supervised (treat) by higher government departments (township, county discipline inspection commission and supervision office)”, “Whether this reform exists in the village this year? (post)”. If the values assigned to both questions are 1, the village is a supervised village, that is, it belongs to the experimental group. If the value given to one is 0, the village is an example of the incomplete realization of supervision over the countryside; that is, it belongs to the control group. The advantage of this scoring is that it can avoid the errors in estimation caused by the strengthening of village democratic supervision or intra-party supervision during the implementation period of the supervision policy. The effect of the policy of assigning supervision to the countryside is related to the integration of intra-party supervision and democratic supervision, rather than the enhancement achieved with one kind of supervisory power alone.$di{d}_{it}$= $trea{t}_i\times pos{t}_t$.

(3)

Control Variable

The factors affecting the effectiveness of public goods supply are very complex. In order to control the influence of other factors over this variable, this paper also sets a series of control variables, including village level and household level.

Village level: The natural characteristics, economic level, political environment and characteristics of village cadres will all have an impact on the construction of rural public goods. With reference to relevant studies [53], here, the total population of the village and the amount of cultivated land are selected as the variables of the village’s natural characteristics. Per capita net income and total income were selected as the variables of village economic level. We chose the distance from the village committee to the nearest township government body and the village chief secretary as political environment variables; the education level, work experience and business experience of village cadres were selected as the characteristic variables of village cadres.

Household level: In the existing research’s results on the factors influencing satisfaction with the supply of public goods, most scholars include a series of variables describing the economic and social characteristics of individuals and family population in the explanatory variables [54]. In this paper, the gender of the household head, household registration, age, education level, being a Party member or not, being in a village cadre or not, and the amount of cultivated land were selected as the control variables at the level of the household of the interviewed farmers.

(4)

Mechanism Variable

The above theoretical analysis shows that the influence of the supervision policy over the effectiveness of rural public goods supply may operate via two mechanisms: the democratic supervisory effect of villagers’ participation, and the restraining effect of village cadres’ behavior. Therefore, in the testing of the mechanism variables, this paper introduces villagers’ participation in the supply of public goods and the amount of investment, as well as the improvement degree of rural education, medical treatment, irrigation facilities and drainage facilities, as proxy variables to measure the effects of villagers’ participation in the supervision, and the restraining effects of village cadres.

(5)

Regulating variable

According to an analysis of the literature, both formal and informal institutions in villages have an impact on the supply of village public goods [11,55]. The implementation of the policy of supervision in the countryside strengthens the power of democratic supervision, as well as the power of intra-party supervision. This will blend with the assessment pressure brought about by the overarching government in the village itself (formal system) and the power of big clan names (informal system), and the influence of the policy will also play a certain regulating role. This paper chooses “Whether the most important assessment standard for village cadres is public service” as a variable to measure the assessment pressure, and selects “Whether the village cadres belong to the clan with the big surname?” as a variable to measure clan power.

The descriptive statistics of the main variables are shown in

Table 2. Measurement methods and descriptive statistical analysis of main variables.

Variables	Measurement Methods	Obs	Mean	Std	Min	Max
Satisfaction with rural education (SRE)	Are farmers satisfied with the public goods supply of this project? 1 = satisfied; 0 = dissatisfied	4637	0.551	0.497	0	1
Satisfaction with rural medical care (SMC)	Same as uplink	8624	0.638	0.480	0	1
Satisfaction with rural irrigation facilities (SIF)	Same as uplink	8451	0.505	0.500	0	1
Satisfaction with rural drainage facilities (SDF)	Same as uplink	6621	0.518	0.500	0	1
Did	Treat × post	8670	0.184	0.388	0	1
Population	Ln (Village population)	8650	7.174	0.685	2.741	9.116
Land	Ln (Village cultivated land)	8570	7.363	1.429	0	9.741
Income	Ln (Per capita income of village)	8650	8.573	0.815	5.303	10.463
Distance	Ln (Distance between the village committee and the nearest town government body)	8631	1.646	0.741	0	4.754
Economy	Ln (Village collective income)	8671	2.300	1.384	0	6.328
Education level of VC	School time of village cadres (year)	8630	3.596	0.815	1	5
Same leader	Whether the village head and secretary are the same person: 1 = yes; 0 = no	8615	0.342	0.474	0	1
Daxing	Whether the village cadres will belong to the big clan: 1 = yes; 0 = no	8642	0.407	0.491	0	1
Workout	Whether the village cadres used to function: 1 = yes; 0 = no	8628	0.168	0.373	0	1
Bussiness	Whether the village cadres do business: 1 = yes; 0 = no	8620	0.269	0.443	0	1
Sex	Gender of household: 1 = male; 0 = female	8668	0.911	0.285	0	1
Hukou	Household registration: 1 = agricultural household; 0 = non-agricultural household	8560	0.966	0.180	0	1
Age	Age of household	8657	54.599	11.097	19	95
Leader	Whether anyone in the family is a village leader:1 = yes; 0 = no	8637	0.091	0.288	0	1
Party	Member of the Communist Party: 1 = yes; 0 = no	8554	0.149	0.356	0	1
Education	School years of household	8645	6.766	3.423	0	24
Cultivated land	Ln (farmland cultivated by households)	8657	1.768	0.904	0	7.098
Comparison of rural education (CRE)	Is it more satisfying than it was three years ago? 1 = yes; 0 = no	4476	0.342	0.474	0	1
Comparison of rural medical care (CMC)	Same as uplink	8476	0.399	0.490	0	1
Comparison of irrigation facilities (CIF)	Same as uplink	8217	0.277	0.448	0	1
Comparison of drainage facilities (CDF)	Same as uplink	6513	0.283	0.450	0	1
Clan constraints	Whether the village cadres hold the biggest surname of the village: 1 = yes (at least one belongs to); 0 = no	8642	0.525	1.499	0	1
Assessment	Is public goods supply the most important assessment indicator? 1 = yes; 0 = no	8671	0.246	0.431	0	1
Ancestral hall	Whether there are ancestral halls in the clan of this village? 1 = yes; 0 = no	8227	0.157	0.365	0	1
Participation	Whether farmers participate in raising funds for village public goods projects? 1 = yes; 0 = no	1918	0.108	0.310	0	1
Raised money	Investment amount of public goods project of household	1733	152.469	1838.305	0	7000

.

3.3. Data Source

The micro-survey data used in this paper came from the China Rural Development Survey, which was conducted between 2005 and 2019 by the Center for Agricultural Policy Research at the Chinese Academy of Sciences.

Members of the research team conducted five field follow-up surveys in the sample villages, and obtained a set of long-term panel data. Sample villages were selected by stratified random sampling. The sampling steps were as follows: First, China was divided into 5 major agro-ecological zones, and a sample province was randomly selected from each agro-ecological zone. A total of 5 sample provinces were selected. These were Jiangsu, Sichuan, Shaanxi, Hebei and Jilin, representing the eastern coastal region, and the southwest, northwest, central and northeast regions in turn. Secondly, in each sample province, all counties were evenly divided into 5 groups according to the per capita gross industrial output value of each county (the counties with the highest 20% per capita gross industrial output value comprised the first group, the second 20% with the next highest industrial output value comprised the second group, and so on). One county was randomly selected from each group as the sample county. Thirdly, in the selected sample counties, each township was divided into two groups in descending order according to the per capita gross industrial output value, and then one township was randomly selected from each group as the sample township. After that, two villages from each township were randomly selected as sample villages via a similar method to that in step three. Finally, 20 households were randomly selected from the sample villages as farmers to be assessed via field survey, and the information of 2000 farmers was collected.

Five rounds of panel data collected via the survey were used in this study. These five surveys inquired about and recorded detailed information on the participation of rural households in the supply of public goods, including the satisfaction degree of different supply projects, the improvement of village public services, the amount of public project financing, the investment intentions of different projects related to public goods, as well as the economic and social characteristics of village population, per capita net income, and village elections, and other detailed information of rural households and families. In addition, the supervision of countryside implementation in the village was assessed through telephone conversations with the village cadres to derive statistics, and this was combined with assessments of the county or township government website’s publicity documents to supplement and improve the sample. Due to the long research interval, missing data, difficulties in follow-up tracking and other reasons, this paper eliminated the samples missing data on the required variables. Finally, the valid samples in this study numbered about 8700. Figure 3 shows the selection of zones.

4. Empirical Results and Analysis

4.1. Baseline Result

The regression results regarding the comprehensive effects of the supervision policy on the supply of rural public goods are shown in

Table 3. (a) Effects of SDC reform on education and medical supplies in rural areas. (b) Baseline: effect of SDC reform to irrigation and drainage facilities in rural areas.

(a)
	Satisfaction with Rural Education (SRE)			Satisfaction with Rural Medical Care (SMC)
Variables	(1)	(2)	(3)	(4)	(5)	(6)
DID	0.025	0.345 ***	0.334 ***	0.091 ***	0.069 ***	0.063 ***
	(0.042)	(0.079)	(0.084)	(0.013)	(0.019)	(0.020)
Income			−0.030			0.063 ***
			(0.023)			(0.013)
Education of VC			0.025 *			0.015 *
			(0.014)			(0.009)
Leader			0.030			0.008
			(0.023)			(0.019)
Education			−0.003			−0.001
			(0.002)			(0.002)
Other control	no	no	yes	no	no	yes
Village FE	no	yes	yes	no	yes	yes
Year FE	no	yes	yes	no	yes	yes
Within R2	0.002	0.072	0.080	0.012	0.021	0.025
Obs	4637	4637	4479	8624	8624	8227
(b)
	Satisfaction with Rural Irrigation Facilities (SIF)			Satisfaction with Rural Drainage Facilities (SDF)
Variables	(7)	(8)	(9)	(10)	(11)	(12)
DID	0.177 ***	0.140 ***	0.134 ***	0.146 ***	0.069 ***	0.073 ***
	(0.014)	(0.0200)	(0.021)	(0.014)	(0.022)	(0.024)
Income			0.073 ***			0.035 **
			(0.013)			(0.016)
Education of VC			0.027 ***			0.012
			(0.009)			(0.011)
Leader			0.061 ***			0.043 *
			(0.019)			(0.023)
Education			−0.005 ***			−0.004 **
			(0.002)			(0.002)
Other control	no	no	yes	no	no	yes
Village FE	no	yes	yes	no	yes	yes
Year FE	no	yes	yes	no	yes	yes
Within R2	0.019	0.022	0.030	0.014	0.024	0.029
Obs	8451	8451	8081	6621	6621	6313

Note: Due to space limitations, the results of other control variables are not listed separately. *, ** and ***, respectively, represent significance at the levels of 1%, 5% and 10%. The following table is the same.

.

In Table 3 (a), columns (1) and (3) list the regression coefficients of policy interaction terms without controlling for the village fixed effect and year fixed effect. Columns (2) and (5) show the regression results when controlling for village fixed effects and year fixed effects, taking only the interaction term of the policy of supervision over the countryside as the explanatory variable. Control variables were further added to columns (3) and (6). The results show that the interaction term of the supervision policy still has a significant positive impact; that is, compared with the non-implementation areas, the policy significantly promotes the degree of satisfaction of rural households with rural education and medical care in the pilot areas.

In Table 3 (b), columns (7) and (10) list the regression coefficients of the policy’s effects on farmers’ satisfaction with the supply of irrigation and drainage facilities, without controlling the village fixed effect and year fixed effect. Columns (8) and (11) report the regression results while controlling for the village fixed effect and year fixed effect, taking only the effect term of the supervision policy as the explanatory variable. Control variables were further added into columns (3) and (6). The results show that the interaction term of the supervision policy still has a significant positive impact; that is, compared with the area of non-implementation, the policy significantly promotes the degree of satisfaction of farmers with irrigation and drainage in the pilot area.

As regards the regression coefficient of the control variables, this paper also found that the higher the per capita net income level of the village, the higher the degree of farmers’ recognition of the effect on village public goods supply, because with a higher per capita income of the village, farmers may have a better understanding of the national social security-related policies related to their own interests than farmers in other areas. Due to the exclusivity and competitiveness of social security public goods, farmers will inevitably mobilize their political resources to obtain relevant benefits and take greater initiative to participate in the supply of public goods. Therefore, farmers in these areas will show a higher level of acceptance and satisfaction with the current situation of the supply of public goods in villages. The educational level of village cadres also has a significant positive impact on the supply effect of rural public goods, because the higher the education level of village cadres, as the main platform of rural economic and social correspondence, the better the implementation of various public goods policies set out by village cadres; rooted in social acquaintance, this can directly connect with and reflect the needs of endogenous public goods construction, and thus more effectively meet the needs of villagers [56].

In addition, the social characteristics of farmers also have a significant impact on the effects of public goods supply. As can be seen from Table 3 (b), if the head of a household is a village leader, this can positively promote villagers’ satisfaction with the supply of public goods. This is because farmers, as village cadres, pay more attention to the situation of the poor and ordinary villagers, which tend to overlap with the groups targeted by rural public goods. Therefore, farmers’ evaluations of the effects on the supply of rural public goods have a significant impact [10]. The higher the educational level of the household’s head, the lower the satisfaction with the effect of mutual assistance on the supply of public goods. This is because the higher the educational level of the household’s head, the higher the demand for village public goods, and the more picky they will be, so they will be more likely to feel dissatisfaction with the current situation. Overall, the empirical results confirm hypothesis H1.

4.2. Heterogeneity

(1)

Level of Economic Development

China’s economy is typical of a big country, and there are great differences in the levels of economic development, factor endowment and industrial development in different regions. In order to further analyze the regional heterogeneity of the optimization of rural public goods supply related to the policy of sending supervision to the countryside, this paper divides villages into economically developed areas and economically underdeveloped areas, according to the average annual GDP rankings of counties given in the National Bureau of Statistics. We then carry out regressions for these two areas. The regression results are shown in Figure 4.

This shows that the supervision policy is significant in areas with weak economic development, and the estimated coefficient is positive but not significant in developed areas. This indicates that, on the whole, this policy can strengthen the effect of improving the supply of rural public goods, but the effect is mainly reflected in regions with weak economic development, namely, the central and western regions. The main reason for this is that the improvements in public goods supply brought about by supervision in the countryside are affected by various factors, such as the completeness of local infrastructure and the level of agricultural development. There is great inequity in China’s rural infrastructural investment, which is mainly caused by inter-regional inequity [51]. Areas with low economic development levels have poor infrastructure and public service conditions. After the implementation of the supervision policy, village cadres will be constrained and will show more room for improvement, and the improvement of supply effects will be more obvious. However, in areas with higher economic development levels, the basic conditions of village public goods are more or less perfect, so the policy’s effect is not significant. Alternatively, the maintenance and management of infrastructure will be neglected because service public goods will be paid more attention, which will lead to a decrease in farmers’ satisfaction with infrastructure.

In short, in areas with lower economic development levels, the improvements in the supply of public goods brought about by supervision will be more significant.

(2)

Poverty-stricken County

Although China had lifted its entire rural population out of poverty by 2020, and all the poor counties have been removed from poverty, significant differences are seen in the work priorities between formerly poor counties and non-poor counties. In this paper, the regions are classified according to whether or not they used to be national-level poverty-stricken counties, and the two types of regions are regressed individually. The results are shown in Figure 5.

It can be seen that the improvement in public goods brought about by the supervision policy obviously differs with regional characteristics and types of public goods. In the villages of national-level poverty-stricken counties, the effect of the supervision policy on the provision of public social services is more significant. According to the regression coefficient, the satisfaction of villagers in poverty-stricken counties with school education increased significantly by 41.93%, and the satisfaction with medical conditions increased significantly by 9.9%. However, in the non-poverty-stricken counties, the promotional effect was not significant. This may be because medical and health services and social security services in poor counties have a prominent effect on poverty reduction; medical and health services can reduce the transient poverty of rural poor families, and education can reduce the transient poverty of families trapped in chronic poverty [57,58]. Therefore, village cadres in poor counties are seen to be more willing to improve education and medical care after greater supervision is given, in order to achieve the poverty reduction target. There are no regional differences in villagers’ satisfaction with irrigation facilities.

Whether in poor counties or non-poor counties, villagers’ satisfaction has been significantly improved, which may be because the national emphasis on agricultural production against the background of the rural revitalization strategy encourages village cadres not to take lightly the management and maintenance of irrigation facilities. Therefore, the satisfaction of villagers with irrigation facilities in all regions has increased significantly. However, the satisfaction of villagers in non-poor counties has increased, which may be because most of the non-poor counties are concentrated in the eastern region (Jiangsu Province). Villages in these areas have sufficient funds for the construction and maintenance of open gullies, dark gullies and shafts, so the improvements in villagers’ satisfaction in these areas were more significant.

4.3. Robustness Check

(1)

Parallel Trend Test

According to the previous study using a differential method, the supervision policy can effectively improve the effectiveness of rural public goods supply. However, a particularly important assumption when using this method is that both the treatment group and the control group satisfy the parallel trend hypothesis before the policy is enacted. In order to verify the reliability of the identification strategy used in the differential method, this paper draws on an existing research framework [59] and uses the event analysis method to investigate the parallel trends and the dynamic effects of the policy in advance. The regression equation is as follows:

$$Y_{ijt}={\sigma }_0+{{\displaystyle \sum }}_{s=-4}^1{\sigma }_{s}di{d}_s+X_{it}^{\prime }\beta +Z_{ijt}^{\prime }\theta +{\gamma }_t+{\phi }_i+{\epsilon }_{ijt}$$

(6)

In model (6), ${\sigma }_s$ is the estimation coefficient, which captures the difference in time trend between “implemented village” and “unimplemented village”. An s value of 0 refers to the implementation year of the pilot policy; an s value that is negative refers to the period before the implementation of the pilot policy; a positive s value means the period after the implementation of the policy. ${\sigma }_0$ is the intercept term, and the remaining variables are consistent with those in model (1).

Figure 6 shows the estimated results of Equation (6). As can be seen from Figure 1, before the implementation of the policy, the estimated coefficients of each time window were not significant. Further, after the implementation of the policy, the coefficient estimates became positive and passed the significance test. The estimation results show that, on the one hand, the treatment group and the control group showed the same change trends before the implementation of the policy, but there was no significant difference; on the other hand, the implementation of the supervision policy in the countryside has had a promoting effect on the effectiveness of the supply of the public goods of school education, medical treatment, irrigation facilities and drainage facilities in rural areas, and the effect of the policy has shown a time lag.

(2)

Placebo Test

Another concern raised about the study’s conclusion is that the statistically significant satisfaction of farmers with the supply of village public goods is likely related to some random factor. Therefore, this paper has employed existing treatment methods in indirect tests [60] to determine whether the positive effect of the supervision policy on the supply of public goods is caused by other non-observed, omitted variables. The samples used in this paper are from 100 villages, among which 56 villages have been supervised. In this paper, 56 villages were randomly selected from 100 villages as fictitious processing groups. Due to the multi-phase differential method used in this paper, multiple time points actually arose when withdrawal and correction occurred. Therefore, after randomly selecting the treatment groups, we further randomly assigned a policy occurrence time point to each village, repeated this 500 times, and then compared the estimated coefficients of the supervised rural policy in the 500 regressions with that achieved in the benchmark regression. Here, the Y-axis represents the density value, the curve is the kernel density distribution of the estimation coefficient, and the vertical solid line is the real estimated value of the DID model. As can be seen from Figure 7, among the 500 regression results generated, most are concentrated near zero, and the p-values of most of the estimated values are greater than 0.1 (not significant at the 10% level), which indicates that our estimated results were unlikely to have been obtained by chance, and therefore unlikely to have been affected by other policies or random factors. Therefore, the probability of “false errors” in this paper is very low.

This further indicates that the supervision policy plays a steady role in improving the effectiveness of rural public goods supply, and indeed improves the degree of satisfaction of villagers with education, medical care, irrigation facilities and drainage facilities.

(3)

Endogeneity: Sample Selection (PSM-DID)

Although relevant tests were conducted on the assumptions and estimated results of the DID method outlined above, self-selection may still have occurred in the group samples. For example, higher-level policy-makers prefer to choose areas with disharmonious relations between cadres and crowds, or areas showing greater modernization in rural governance, as pilot areas, thus leading to sample selection bias. Therefore, in order to reduce the effects of sample self-selection on the regression results, this paper further used the propensity score-matched multiple difference method (PSM-DID) to test the effects of the rural supervision policy.

In this paper, the Logit model has first been used to calculate a propensity score of each village area subjected to the rural supervision policy. The covariables in the model include village economic status variables, natural condition variables, social security degree variables and village level control variables, which do not include the characteristics of the village cadres mentioned above. Then, according to the propensity score, a sample-matching control group of supervised rural reform areas has been selected. The explanatory variables affecting the two groups of samples were not significantly different before the policy was enacted, so as to reduce sample selection bias. In this study, the kernel matching method was used for matching. In order to ensure the matching quality, the balance test and common support test were carried out at the same time. The relevant test results are shown in Figure 8 and Figure 9. It can be seen that the standardization deviation of most variables was large before matching, while that of all variables was greatly reduced after matching, and all were less than 10%. This indicates that there is no systematic difference between the treatment group and the control group, which meets the requirements of a randomized experiment. In addition, the samples basically satisfy the co-supporting hypothesis. On this basis, samples outside of the common support domain were eliminated, and the DID method was used to estimate the net effect of the supervision policy on the effectiveness of rural public goods supply. Since the PSM method can be used to solve the sample selection bias problem most effectively, and the DID method can eliminate the influence of missing variables, a combination of the two methods can be used to effectively identify the improvement and promotion effects of the policy.

The regression results of the PSM-DID method are shown in

Table 4. Robustness check: matching method I and II.

(1) Robustness Check: Matching Method I
	Satisfaction with Rural Education (SRE)			Satisfaction with Rural Medical Care (SMC)
	Nearest Neighbor Matching	Nuclear Matching	Nuclear Matching	Nearest Neighbor Matching	Nuclear Matching	Nuclear Matching
Variables	(1)	(2)	(3)	(4)	(5)	(6)
DID	0.333 ***	0.337 ***	0.335 ***	0.056 ***	0.059 ***	0.041 *
	(0.089)	(0.090)	(0.091)	(0.022)	(0.021)	(0.024)
Control	yes	yes	yes	yes	yes	yes
Village FE	yes	yes	yes	yes	yes	yes
Year FE	yes	yes	yes	yes	yes	yes
within R2	0.076	0.071	0.081	0.021	0.020	0.020
obs	3913	3974	3438	7325	7426	6335
(2) Robustness Check: Matching Method II
	Satisfaction with Rural Irrigation Facilities (SIF)			Satisfaction with Rural Drainage Facilities (SDF)
	Nearest Neighbor Matching	Nuclear Matching	Nuclear Matching	Nearest Neighbor Matching	Nuclear Matching	Nuclear Matching
Variables	(7)	(8)	(9)	(10)	(11)	(12)
Did	0.151 ***	0.144 ***	0.143 ***	0.081 ***	0.070 ***	0.096 ***
	(0.022)	(0.022)	(0.025)	(0.025)	(0.025)	(0.028)
Control	yes	yes	yes	yes	yes	yes
Village FE	yes	yes	yes	yes	yes	yes
Year FE	yes	yes	yes	yes	yes	yes
within R2	0.029	0.028	0.027	0.027	0.026	0.029
obs	7199	7300	6225	5665	5745	4957

Note: Due to space limitations, the results of other control variables are not listed separately. * and ***, respectively, represent significance at the levels of 1% and 10%.

. The results show that the core explanatory variables are significant, and the estimated coefficients are positive, regardless of the matching method used. Therefore, the hypothesized positive influence of the supervision policy in improving the public goods supply effect is robust.

(4)

Other Robustness Test

First, as regards the implementation of the supervision policy in the countryside, according to the official announcements collected from county and township governments, the pilots began in 2009. In the sample used in this paper, this reform had been implemented in four villages in 2008, which figure may result from the different calibers of inquiry used during the surveys. In order to ensure the reliability of the regression results, the villages in 2008 were excluded from the sample before the regression was carried out. The specific test results are shown in the even-numbered columns of

Table 5. Robustness check: remove samples I and II.

(1) Robustness Check: Remove Samples I
	Satisfaction with Rural Education (SRE)		Satisfaction with Rural Medical Care (SMC)
	Excluding 2005	Remove Samples in 2008	Excluding 2005	Remove Samples in 2008
Variables	(1)	(2)	(3)	(4)
DID	0.273 **	0.546 ***	0.037 *	0.042 **
	(0.129)	(0.093)	(0.021)	(0.021)
Control	yes	yes	yes	yes
Village FE	yes	yes	yes	yes
Year FE	yes	yes	yes	yes
within R2	0.101	0.080	0.028	0.024
obs	3724	4291	7471	7911
(2) Robustness Check: Remove samples II
	Satisfaction with Rural Irrigation Facilities (SIF)		Satisfaction with Rural Drainage Facilities (SDF)
	Excluding 2005	Remove Samples in 2008	Excluding 2005	Remove Samples in 2008
Variables	(5)	(6)	(7)	(8)
DID	0.143 ***	0.139 ***	0.078 ***	0.062 **
	(0.022)	(0.022)	(0.026)	(0.025)
Control	yes	yes	yes	yes
Village FE	yes	yes	yes	yes
Year FE	yes	yes	yes	yes
within R2	0.030	0.031	0.022	0.028
obs	7386	7764	5629	6072

Note: Due to space limitations, the results of other control variables are not listed separately. *, ** and ***, respectively, represent significance at the levels of 1%, 5% and 10%.

. It can be seen that the implementation of the supervision policy has significantly improved the effectiveness of the supply of village public goods, and the degree of satisfaction with the four selected categories of public goods is significantly positive, which again verifies the robustness of our research conclusion.

Second, after the completion of tax and fee reforms in 2006, a large number of resources were sent to rural areas, and a “project system” soon became the main mode of rural public goods supply. In order to exclude the effects of tax reform on the estimated results, this paper has excluded survey data from 2005, and selected the four survey years of 2008, 2012, 2016 and 2019 for regression. The specific test results are shown in the columns of Table 5. It can be seen that the interaction term of the rural supervision policy is still significant, and the estimated coefficient is positive, which indicates that our research conclusion is robust.

Third, since the dependent variable selected in this paper is a binary categorical variable of 0 or 1, we have also chosen to replace the Logit model for the probit model in the robustness test. The specific test results are shown in

Table 6. Robustness check: change empirical model.

	SRE	SMC	SIF	SDF
	Logit	Logit	Logit	Logit
Variables	(1)	(2)	(3)	(4)
DID	0.2941	0.067 ***	0.130 ***	0.071 ***
	(3.133)	(0.021)	(0.021)	(0.024)
Control	yes	yes	yes	yes
Village FE	yes	yes	yes	yes
Year FE	yes	yes	yes	yes
Log pseudolikelihood	−2494.096	−4873.057	−4726.515	−3944.372
Wald chi2	538.240	770.560	1187.970	615.190
Observations	4451	8227	8081	6313

Note: Due to space limitations, the results of other control variables are not listed separately. ***, represent significance at the levels of 10%.

. It can be seen that, in addition to the school education coefficient, which is not significant, the effectiveness values of the other three types of public goods supply are still significant, and the estimated coefficient is positive, which indicates that our research conclusion is basically robust.

5. Mechanism Test

5.1. Trade-Off

The above regression results verify that the supervision policy will significantly improve villagers’ satisfaction with the provision of public goods, but the question remains: what is the intermediate mechanism of this policy, by which it improves the effectiveness of the provision of public goods? Therefore, combined with the theoretical analysis above, we have introduced the relevant variables of villagers’ public participation and the proxy variables of the constraining effects of village cadres in order to explore the mechanism of the effect of the supervision policy on the supply of village public goods. Due to the lack of data, the data for 2012 have been selected for verification in this paper.

The above analysis is based on Formula (1). After the implementation of the policy of directing supervision towards the countryside, the discipline inspection commissioners of each village come under the direct jurisdiction of, and are subjected to annual evaluations by, the township and county governments, which strengthens the checks and balances of the discipline inspection commissioners as regards the cadres of the village committee, as well as the supervisory powers of the village discipline inspection commissioners, who also serve as the directors of the village supervision committees, while further increasing the number channels by which villagers can report to the Party organization, the village cadres and the overarching government. There are two kinds of constraining effects: villager participation and higher government intervention. For this reason, we have selected the willingness of villagers to participate in the supply of public goods to conduct a regression analysis. The test results are shown in

Table 7. Mechanism test: household participation and SDC reform.

	Participation in Investment	Investment Amount
Variables	(1)	(2)
DID	0.771 **	897.3 **
	(0.349)	(357.2)
Control	yes	yes
Village FE	yes	yes
Year FE	yes	yes
R2	0.282	0.060
Observations	1890	1733

Note: Due to space limitations, the results of other control variables are not listed separately. **, represent significance at the levels of 5%.

and

Table 8. Mechanism test: household participation and effect on public goods supply.

	SIF	SDF	SRE	SMC
Variables	(1)	(2)	(3)	(4)
Trade-off one
Participation in investment	0.069 *	0.085 *	0.074 *	−0.031
	(0.041)	(0.048)	(0.044)	(0.044)
Control	yes	yes	yes	yes
Village FE	yes	yes	yes	yes
Year FE	yes	yes	yes	yes
R2	0.336	0.191	0.236	0.254
Observations	1437	1235	1290	1290
Trade-off two
Investment amount	0.022 *	0.021 *	0.021 *	0.006 **
	(0.012)	(0.013)	(0.012)	(0.002)
Control	yes	yes	yes	yes
Village FE	yes	yes	yes	yes
Year FE	yes	yes	yes	yes
R2	0.306	0.182	0.237	0.217
Observations	1707	1558	1518	1620

Note: In order to make the data more intuitive to read, the unit of investment amount in this verification is converted to “CNY 1000”. The meaning of the coefficient is the probability that the corresponding degree of satisfaction will change with each increase of CNY 1000. Due to space limitations, the results of other control variables are not listed separately. * and **, respectively, represent significance at the levels of 1% and 5%.

.

Columns (1) and (2) in Table 7 present the effects of villagers’ public participation as induced by the supervision policy. The results show that the pilot policy significantly improves the level of local villagers’ participation in the supply of public goods, that is, after the implementation of the supervision policy, the probability of villagers’ willingness to participate in the investment activities of public goods projects will be significantly increased by about 77.1%. Moreover, the contribution that rural households are willing to make will increase significantly by about CNY 897, and this reaches the significance level of 5%, which is consistent with our expectations. This is because, after supervision is directed towards the countryside, the villagers’ rights to participate in public affairs and make suggestions to the village-level organization are further guaranteed, which lays the foundations for strengthening supervision within the village. Moreover, these facilities are also daily necessities for villagers who have lived in the countryside for a long time, so the villagers’ enthusiasm to participate in public goods supply projects will be enhanced.

According to the results shown in Table 8, the more villagers participate in investment, and the higher the amount of investment, that is, the deeper the degree of villagers’ participation, the more obvious the improvement in rural satisfaction with public goods will be; the coefficient is significantly positive at the level of 10%. Combined with the results in Table 8 and

Table 9. Mechanism test: behavior of village cadres and SDC reform.

	Is This Public Goods Project Any More Satisfying/Better than It Was Three Years Ago?
	Education	Medical Care	Irrigation Facilities	Drainage Facilities
Variables	(1)	(2)	(3)	(4)
DID	0.287 ***	−0.033	0.054 **	0.065 ***
	(0.095)	(0.022)	(0.021)	(0.023)
Control	yes	yes	yes	yes
Village FE	yes	yes	yes	yes
Year FE	yes	yes	yes	yes
within R2	0.205	0.061	0.040	0.031
Observations	3981	8087	7223	5724

Note: Due to space limitations, the results of other control variables are not listed separately. ** and ***, respectively, represent significance at the levels of 5% and 10%. The following table is the same.

, it can be seen that supervision in the countryside can indeed promote the level of villagers’ participation in public affairs, and thus improve the effectiveness of public goods supply.

The higher levels of government will strengthen the intra-party supervisory power, which will increase the pressure placed on village cadres in completing tasks, and this supervisory effect will force village cadres to focus more on completing their duties. With the strengthening of the supervisory effect over villagers, village cadres will come to be supervised by both external and internal forces, so they will be more cautious in handling daily public affairs, which is also reflected in our results. In this paper, “The comparison of villagers’ supply conditions of various public goods” is selected as the test variable for regression, to test whether the present situation regarding the supply of public goods in villages is significantly improved after the implementation of the policy of directing supervision towards the countryside.

According to the results shown in Table 9, medical projects aside, all villagers surveyed thought that the education, irrigation and drainage facilities had been significantly improved, which indicates that after supervision is directed towards the countryside, the village cadres’ work behavior becomes more restricted, and they take on more responsibilities and obligations in devoting themselves to public affairs and the service of the village. In order to test this mechanism, we have also taken the villagers’ satisfaction with the effectiveness of public goods provision as the dependent variable, and put “before and after comparison of the villagers’ supply of various public goods” into the regression model. The results are shown in

Table 10. Mechanism test: behavior of village cadres and effect on public goods supply.

	Satisfaction with the Supply of Public Goods
	Rural Education	Rural Medical Care	Irrigation Facilities	Drainage Facilities
Variables	(1)	(2)	(3)	(4)
Comparison of rural education (CRE)	0.408 ***
	(0.021)
Comparison of rural medical care (CMC)		0.512 ***
		(0.017)
Comparison of irrigation facilities (CIF)			0.408 ***
			(0.016)
Comparison of drainage facilities (CDF)				0.446 ***
				(0.016)
Control	yes	yes	yes	yes
Village FE	yes	yes	yes	yes
Year FE	yes	yes	yes	yes
within R2	0.140	0.097	0.066	0.073
Observations	4419	8155	7996	6295

Note: Due to space limitations, the results of other control variables are not listed separately. *** represent significance at the levels 10%.

. As can be seen from the results in Table 10, the corresponding coefficients influencing all kinds of public goods are significantly positive at the level of 1%, which is consistent with expectations, so the effect of the supply of public goods in villages will be significantly improved using this approach.

The above estimation results verify hypothesis H2 in this paper.

5.2. Regulating Effect I: Coupling of Clan Power and Village Authority

In village governance, the interaction between clan and formal systems will affect the supply of the village’s public goods. Some studies have revealed the restraining effect of the clan on the formal authority of the village in seeking to achieve the effect of promoting the supply of public goods, that is, the restriction of village cadres by clan organizations’ in the village will significantly increase the supply of public goods in the village [34,39]. However, other studies have found the exact opposite effect. Their research found that a village director from a powerful clan would, on the contrary, lead to the greater supply of private goods, and reduce the supply of public goods [49]. Therefore, the clan forces do have a certain inhibiting and restraining effect on the formal authority of the village. However, when this influence is confronted with the formal system of the inner-party’s supervisory power, the impact on the supply of public goods remains to be discussed.

This paper draws on previous studies, adopts the top three surnames in the village to substantiate the “clan with the big surname”, and matches the surnames of the village cadres with the big surnames of the village [61,62]. In the questionnaire, we used the question “Are the village cadres and their spouses from the big surname family?” as a variable to measure the infiltration of clan forces into the formal authority of the village, and the interaction term of clan forces and the countryside supervisory policy is crossed into the model for the regression analysis. In the process of the election of village cadres, villagers tend to elect members of the same ethnic group. The fact that village cadres tend to come from big families can be used to infer the degree of clan network penetration of and interference in village governance, to a certain extent, allowing us to better measure the strength of the clan network. The regression results are shown in

Table 11. Regulating effect: clan power with SDC reform.

	SRE		SMC		SIF		SDF
Variables	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)
DID	0.335 ***	0.162	0.064 ***	0.100 ***	0.136 ***	0.166 ***	0.070 ***	0.078 **
	(0.084)	(0.173)	(0.020)	(0.025)	(0.021)	(0.027)	(0.024)	(0.029)
Clan constraints	0.052	0.042	−0.024	−0.012	−0.026	−0.015	0.041	0.043
	(0.039)	(0.040)	(0.022)	(0.022)	(0.023)	(0.023)	(0.029)	(0.029)
DID × Clan constraints		0.197		−0.069 **		−0.058 *		−0.015
		(0.150)		(0.030)		(0.030)		(0.035)
Control	yes	yes	yes	yes	yes	yes	yes	yes
Village FE	yes	yes	yes	yes	yes	yes	yes	yes
Year FE	yes	yes	yes	yes	yes	yes	yes	yes
within R2	0.080	0.081	0.025	0.026	0.030	0.031	0.029	0.030
obs	4479	4479	8227	8227	8081	8081	6313	6313

Note: Due to space limitations, the results of other control variables are not listed separately. *, ** and ***, respectively, represent significance at the levels of 1%, 5% and 10%. The following table is the same.

.

As can be seen from columns (3) to (6) in Table 11, the interaction coefficient between the policy interaction terms and the village cadres’ big surnames is negative and significant, indicating that the effects of clans with big surnames on village formal authority weaken or inhibit the improvement in the public goods supply of medical and irrigation facilities brought about by the supervision policy, and a significant negative regulating effect is shown. This may be due to the reduction in clan power in public organizations and the mobilization resulting from the mass outflow of labor force. At this time, village cadres understand that the demand for public goods may not be consistent with the demands of most villagers [63]. Villagers seek improved medical and irrigation facilities more urgently, but village cadres pay more attention to the governance of the village environment and other aspects. Then, after the implementation of the supervision policy, the improvements in the supply of public goods were weakened in villages where the cadres were from a big clan. This also confirms hypothesis H3.

The intensity of the clan network will reflect the penetration of the clan network’s influence. In order to further study the actual role of clan penetration intensity in the process of supervision as applied to the countryside, this paper reflects the organization and cohesion of the clan network via indicators such as “whether there are ancestral temples”, based on the model settings of existing research [34]. The influence of clans with ancestral temples in their villages is greater than that of clans from villages without ancestral temples. Therefore, the public goods of medical and irrigation facilities, with significant moderating interaction terms (Table 11), are selected as the dependent variable in this paper, and “with or without ancestral temples” is taken as the classification basis to judge the strength of clan power. The results are shown in

Table 12. Regulating effect: clan power with SDC reform in groups.

	Rural Medical Care		Irrigation Facilities
	Strong Clan Power	Weak Clan Power	Strong Clan Power	Weak Clan Power
Variables	(1)	(2)	(3)	(4)
Did	0.183 ***	0.0958 ***	0.349 ***	0.145 ***
	(0.067)	(0.028)	(0.066)	(0.030)
Clan constraints	−0.036	−0.011	0.035	−0.016
	(0.058)	(0.025)	(0.059)	(0.026)
Did × Clan constraints	−0.101	−0.068 **	−0.025	−0.078 **
	(0.075)	(0.034)	(0.070)	(0.034)
Control	yes	yes	yes	yes
Village FE	yes	yes	yes	yes
Year FE	yes	yes	yes	yes
within R2	0.106	0.119	0.203	0.202
obs	1295	6932	1280	6801

Note: Due to space limitations, the results of other control variables are not listed separately. ** and ***, respectively, represent significance at the levels of 5% and 10%.

. It can be seen from the results in the table that the interaction coefficient of the regulation effect in areas with weak clan influence is negative and significant at the level of 5%, and the influence coefficient of supervision in the countryside is significantly positive. However, the influence coefficient of the supervision policy in areas with strong clan power is significantly positive, and about 2.5 times greater than that in areas with weak clan power. This shows that in areas with strong clan power, the communication and restraining effects between the big clan members can cause village cadres of specific ethnicities to pay attention to the needs of villagers actively, and provide necessary public goods. However, in villages with weak clan power, there is tension between the supervision provided by the clan and the collective action function. At this time, village cadres of the “Qiangzong Da” nationality tend to lack incentives and internal supervision, and collective inaction and moral hazards may arise [48,62], so negative emotions and negative effects inevitably arise after they move to the countryside under the effects of supervision.

This study’s results also reflect that informal institutional force (clan) penetrates into the formal authority of the village, and the coupling between the informal institutional force (clan supervision) and the formal institutional supervision force (strengthened intra-party supervision) will not necessarily induce an ideal strengthening effect. The two need to be perfectly coordinated, and the clan’s internal supervision and moral restraint functions need to be brought into play.

5.3. Regulating Effect II: Village Cadre Assessment

As the main vehicle of village governance, village cadres are not only supervised by the informal system inside the village (such as the village group, and the clan with the biggest surname), but are also constrained by the external formal system (such as assessment by higher levels of government, such as the township government). On the one hand, some external formal institutions will significantly promote the supply of public goods. Existing studies have found that a grassroots democratic system can significantly enhance the importance of public goods to rich village cadres, and thus improve the supply level of public goods [64], and the implementation of a “shoulder to shoulder” system in village cadres significantly promotes the supply of village collective irrigation facilities [65]. On the other hand, if the external institutional constraints are too strong, the responsibilities of village cadres will become administrative; that is, village cadres will become a communicational link between superiors and farmers, and village cadres will lose the ability to organize and mobilize farmers to form demands related their common interests; further, collective and village cadres will lose their ability to influence village public opinion, and this may even result in the disappearance of village publicity [18]. Therefore, the influence of external formal systems on village cadres’ governance and their provision of public goods is not clear. Therefore, according to the differences between the assessment indicators of cadres in each village, this paper selects “whether the most important assessment indicator of village cadres in this year is public service” as the proxy variable of how external formal systems supervise or constrain village cadres, and crosses it with the interaction term of the supervision policy applied to the countryside in the model of regression analysis. The test results are shown in

Table 13. Regulating effect: assessment of village cadres with SDC reform in groups.

	Satisfaction with Rural Education (SRE)		Satisfaction with Rural Medical Care (SMC)		Satisfaction with Rural Irrigation Facilities (SIF)		Satisfaction with Rural Drainage Facilities (SDF)
Variables	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)
DID	0.284 ***	0.283 **	0.066 ***	0.098 ***	0.135 ***	0.161 ***	0.065 **	0.103 ***
	(0.105)	(0.123)	(0.022)	(0.027)	(0.023)	(0.024)	(0.026)	(0.028)
Assessment	0.069 **	0.069 **	0.017	0.042 **	0.015	0.035 **	0.026	0.062 ***
	(0.029)	(0.029)	(0.015)	(0.017)	(0.015)	(0.016)	(0.018)	(0.022)
DID × Assessment		0.002		−0.116 ***		−0.095 ***		−0.134 ***
		(0.133)		(0.034)		(0.035)		(0.042)
Control	yes	yes	yes	yes	yes	yes	yes	yes
HH FE	yes	yes	yes	yes	yes	yes	yes	yes
Year FE	yes	yes	yes	yes	yes	yes	yes	yes
within R2	0.086	0.086	0.026	0.028	0.031	0.032	0.032	0.035
obs	4479	4479	8227	8227	8081	8081	6313	6313

Note: Due to space limitations, the results of other control variables are not listed separately. ** and ***, respectively, represent significance at the levels of 5% and 10%.

.

In this paper, the variables of the village cadre assessment index are assessed on the basis of Formula (1), and the regression results are shown in the list below. We have then added the policy interaction and cross-multiplication items of the village cadre assessment on this basis. The regression results are shown in the even-numbered columns in the table below. From the cross-multiplication coefficients of columns (4), (6) and (8) in Table 13, it can be seen that the supervisory power brought about by the village cadre assessment mechanism of the formal system weakens or inhibits the improvement effect of the supervision policy on the supply of medical, irrigation and drainage facilities; that is, it has a significant negative regulating effect. However, the assessment of village cadres (the moderating variable) itself has a significant effect on the improvement of the supply of public goods, which indicates that the restraining effect of “assessment of village cadres by the superior government” present in each village, and the restraining effect of “strengthening intra-party supervision”, brought about by the implementation of the supervision policy in the countryside, are relations that can obviously be substituted. This can also explain why the effectiveness of the supply of public goods will decline under the restrictions of the evaluation and supervision of village cadres in the countryside. In the process of duty fulfillment, village cadres should play the role of “benefiting the village” on behalf of the government and villagers to safeguard the practical interests of both sides. However, due to the psychological and behavioral motivations of individuals, village cadres tend to take the maximization of their own interests as the start and end points of all their behaviors, which leads to a tendency towards the “self-interested disadvantaging of the village” in their behavior [66]. Therefore, when the constraints of external formal system are imposed on village cadres at the same time, the village cadres are more likely to display a rebellious tendency due to the “egoistic” impulse, which reduces the efficiency of village governance. This verifies hypothesis H4.

To sum up, the two kinds of supervision force, the assessment of village cadres and the supervision of the countryside, belong to formal systemic constraints; they can be replaced by each other, and the excessive pressure of the two superimposed will lead to a reduction in the supply effect, which also confirms the principle that supervision must be applied to an appropriate degree, otherwise there will be “too much of a good thing”.

6. Conclusions

In order to investigate the relationship between village-level power supervision and the supply of public goods in rural areas, and to accurately assess the actual impact of the supervision policy on the supply of public goods, this paper adopts the panel data of 100 villages in five provinces from 2005 to 2019, and undertakes a large number of systematic analyses using the PSM-DID differential method and the partial method of tendency matching. The study found the following.

First, the rural supervision policy has significantly improved the supply of public goods in villages, and increased the satisfaction of villagers with their economic basis and the social services provided by public goods. Second, there is regional heterogeneity in the positive effect of the policy on the supply of public goods, which is mainly reflected in areas with weak economic development levels. Formerly poor counties experience more significant effects on the social services of public goods, while non-poor counties experience more significant effects on basic economic public goods. Thirdly, this paper also finds that the policy mainly promotes the optimization of public goods supply through mechanisms such as strengthening villagers’ participation in supervision, and the restraining effects of village cadres’ behavior. Fourthly, as regards the regulating effect of clan power, this paper finds that clan power penetrates into the formal authority of the village, and its coupling with the formal institutional supervisory power does not necessarily lead to an ideal strengthening effect. The influence of clan power in the village also needs to be kept strong enough that the actual demands for collective action levied on the villagers are consistent. Besides this, it is found that the force of formal supervision originating from the same source can be substituted for the force of village cadres.

7. Discussion

In terms of the current degree of implementation of supervision in rural areas, and its impact on the development of rural areas, the research findings of this paper have multiple policy implications.

Firstly, the research conclusions of this paper provide an important practical basis from which the government can further support the reform of grassroots discipline inspection and procuratorial work. The research results of this paper show that the government can delegate grassroots inner-party supervision to the village level, and this can improve the effectiveness of rural public goods supply to a certain extent, and enhance the satisfaction of villagers. Therefore, it is necessary to extend the pilot testing of rural areas supervision to a larger scale, such that the results of the pilot test will benefit more farmers and regions.

Secondly, the research conclusions of this paper also provide a useful reference in preventing excessive government supervision or intervention from causing village administration and village publicity losses. Under the supervisory policy, villages should give a full guiding role to the clan with the biggest surnames. Under the premise of ensuring that the interests of the clan with big surnames are consistent with the interests of most villagers, they should undertake democratic supervision and organize collective actions, so as to more effectively guarantee the supply of rural public goods. In addition, the force of the countryside supervision policy and the restraining effect brought about by the evaluation index of village cadres substitute each other, such that the superior government should relax the evaluation standards of village cadres in order to strengthen supervision at the village level, and reduce the immoral behaviors of village cadres via reverse psychology, as these reduce the efficiency of village governance.

The contribution of this paper lies in adding the research angle of the influence of village-level power supervision over rural public goods, and the finding that there is an “inflection point” in the coupling between formal and informal institutions in rural society. It is necessary to improve the efficiency of the interaction between government departments and clan networks, so as to enhance the positive effects of superposition and weaken the possible negative effects. In addition, exogenous supervisory powers can also form a substitutional relationship with the original supervisory institution of the village, offering new empirical evidence for the view that power supervision at the village level is limited.

The deficiency of this paper is that it does not represent the actual attitudes and views of village cadres after the implementation of supervision in the countryside, and these cannot be analyzed and tested with reference to the specific behaviors of the village cadres. Future research avenues include: (1) Analyzing the influence of different degrees of supervision on the actual public goods supply behavior of village committees and village cadres. (2) Analyzing relevant policies that will directly affect the village-level power supervision system.