Effects of the Supervision Down to the Countryside on Public Spending: Empirical Evidence from Rural China

Abstract

Improving the supply of rural public goods serves as a driving engine for rural revitalization and provides fundamental assurance for achieving self-sustained development in rural areas. This study examines how China’s supervision down to the countryside (SDC) policy affects village-level public expenditure, addressing broader debates on grassroots governance reforms. Using 2005–2019 panel data from 100 villages across five provinces, we employ a multi-period staggered difference-in-differences (DID) design to identify causal effects. Empirical results indicate that SDC implementation significantly reduced overall village public spending and investment in new public goods, primarily driven by enhanced budget constraints. Case analysis reveals that this occurs through procedural formalization and participatory oversight. Heterogeneity analysis shows that the effect is more pronounced in villages with weaker clan influence, lower economic development, and absence of factional competition. The findings of this study provide empirical evidence for the perspective in village power supervision theory that “top-down, external, institutional supervision requires clearly defined boundaries” and provides a reference for policies aimed at promoting the sustainable development of rural governance.

1. Introduction

Fundamentally, for the realization of rural revitalization, constructing a comprehensive and modern rural public infrastructure that reaches villages and extends to households is of the utmost importance. However, farmer expectations exceed actual infrastructure and public service delivery. There exist numerous institutional issues, such as significant regional disparities in supply, inefficiencies in public goods provision, and imbalances in public goods investment and spending [1,2,3]. It is of great value to study how to maintain rural public investment and improve the quality of rural public goods under the current system, and lay the foundation for the integration of urban and rural services.

Study of rural public goods supply in China has always been a focal point in the academic realm. From the perspective of the subjects and methods of the supply of rural public goods, the supply of village public goods mainly includes four modes: government supply, market supply, supply by village self-governance organizations, and individual farmer supply. It is generally believed in the academic community that government supply is the most stable mode. However, in actual implementation, it has exposed institutional problems such as the coexistence of over-supply and under-supply, mismatches between supply and household demand, and the involution of grassroots governance caused by the resource-sharing order [4,5,6]. Consequently, the role of village-level organizations in public goods supply has drawn increasing attention.

Enhancing grassroots supervision is a crucial measure for optimizing the supply of rural public goods and facilitating rural revitalization. With the extension of Party-internal supervision to supervision at village level, aiming to achieve “last-mile supervision”, the grassroots supervision system has been gradually systematized.

Some research indicates that when state power infiltrates the village level through various means, it may lead to the bureaucratization of village cadres and the semi-bureaucratization of self-governance for villages. This results in the involution of village funds, an increase in village governance costs, and organizational challenges in public goods supply [7]. Conversely, other studies suggested that social organizations could endow the public with the rights to information and supervision, guarantee villagers’ participation, establish a supply model with mass-participation, and implement grassroots supervision mechanisms [8,9,10]. Although current research on public goods supply has focused on the impact of village elites like village cadres, it has overlooked the restrictive effect of township government supervision on village cadres, and it has not conducted a detailed analysis and empirical test of the internal mechanism regarding the impact of this top-down formal supervision on village public governance. This study takes the supply of rural public goods as the research object and analyzes the impact of supervision policies on village governance.

This study tries to bridge the gap regarding research related to supervision and rural public goods supply. Supervision down to the countryside (SDC) represents a reform of the power supervision system in villages. The sociological field posits that the State first constructs a complete inspection, supervision, and accountability system at the township grassroots. Subsequently, this supervision mechanism is extended to village-level organizations, imposing new requirements on village cadres for village governance and serving villagers, with the aim of achieving “last-mile supervision” [11]. In the existing literature, research directly analyzing the relationship between village-level supervision and the participation of village committees in public goods supply is scarce.

The research findings can be broadly categorized into two viewpoints. Some argued that village-level supervision strengthens the influence of higher-level governments in village governance but weakens governance effectiveness at the grass-roots level. It transforms the process of resource and project allocation to villages into an interest-distribution process [12,13]. The entry of state power at village level may lead to the bureaucratization of village cadres and the semi-bureaucratization of village governance, giving rise to grassroots formalism [7,11]. Others believe that village supervision standardizes the official conduct of village cadres, and forms a comprehensive management approach, which may potentially promote public goods supply [14].

Although the above-mentioned literature roughly analyzed the possible impact of village-level supervision on the supply of public goods by village committees through some village cases, the academic community’s conclusion on this impact is not uniform. Moreover, existing studies have insufficiently analyzed the internal transmission mechanism of this influence. Although some scholars have found that mechanisms such as interest-driven and risk avoidance can lead to the failure of village governance, they have overlooked the fact that process standardization, the pressure of cadre assessment, and village customs can play a regulatory role in this. Due to data issues, few studies have been able to analyze using empirical testing.

Compared with previous studies, this paper makes the following three possible contributions. Firstly, by utilizing the panel data of the China Rural Development Survey (conducted in 2005, 2008, 2012, 2016, and 2019) and taking the supervision policy (SDC) as a quasi-natural experiment, this paper examines the impact of village-level supervision on the public goods supply behavior of village committees. By applying the DID method, it overcomes the estimation biases in previous research, enhancing the reliability of the regression results and providing empirical evidence for existing viewpoints. Secondly, the paper not only conducts a detailed analysis of the short-term weakening effect and regional heterogeneity of the village-level supervision reform on the public goods supply by village committees, but also discovers that the assessment criteria for village cadres can mitigate the negative impacts of the policy, deepening our understanding of the policy’s weakening effect on public goods supply. Finally, it explores the internal transmission mechanism through which the supervision policy restricts the public goods supply via village committees, from the perspectives of cost-cutting and reduced investment in village public goods. This provides practical countermeasures and suggestions for further improving the village-level supervision policy and promoting rural modernization.

The paper is structured as follows: In Section 2, it presents the policy background and theoretical analysis; in Section 3, it introduces the data and model; in Section 4, it focuses on the estimation results and analysis, covering benchmark regression results, heterogeneity analysis, and robustness tests; in Section 5, it explores the mechanism; and the conclusions and policy implications are provided in Section 6.

2. Policy Background and Theoretical Analysis

2.1. Rural Supervision in China

Villager autonomy constitutes the endogenous mechanism for public power supervision in Chinese villages. Village public power is collectively delegated by all villagers, with the village committee (VC) managing public affairs as their agent. The VC is defined as a mass autonomous organization: neither a political body nor a political party entity, though it operates under the leadership of village Party organizations. Villagers’ self-governance embodies grassroots democratic practice through “democratic elections, decision-making, management, and supervision.”

Two additional organizations operate at the village level: the village Party committee (VPC) and the village supervision committee (VSC). The VSC was established with members selected from petitioning households or highly respected villagers to institutionalize interorganizational power balance. The original supervision model positioned the VPC and VSC as representatives of higher-level governments and villagers, respectively, tasked with overseeing VC members, coordinating economic development with village governance, and serving farming households (Figure 1).

However, this supervision system proves ineffective due to two primary factors: (1) Deficiencies in institutional design. The village supervision committee (VSC) primarily comprises village residents, while all internal and administrative affairs remain under the village committee’s (VC) jurisdiction. Coupled with China’s rural communities operating as relationship-based societies, these conditions compromise the VSC’s independence. Additionally, the absence of direct linkages between the VSC and higher-level governments results in weak institutional incentives for proactive oversight. (2) Villager capacity constraints. Villagers demonstrate limited awareness of autonomous supervision and possess inadequate capabilities for effective monitoring.

The supervision down to the countryside (SDC) policy addresses ineffective oversight by establishing a vertical management system where county/township committees integrate external supervision into village governance.

This institutional redesign includes three core components: First, township discipline commissions appoint village discipline inspectors who concurrently serve as VSC chairs and integrity supervisors, directly reporting to township authorities (monthly village briefings; biannual township reports), with direct county-level reporting channels for critical livelihood issues (Figure 2). Second, inspector incentives are strengthened through township-funded stipend supplements to county-level allowances, coupled with performance-linked bonuses integrated into annual leadership evaluations. Third, collective affairs undergo procedural standardization via codified sequences—proposal, deliberation, decision, audit, implementation, and supervision—enforcing pre-implementation villager consultation, mid-implementation public disclosure, and post-implementation outcome evaluations.

The redesign improved the existing internal supervision mechanism and achieved the “last mile” of supervision.

The local government issued an official document emphasizing that “the annual assessment of the village discipline inspection committee is responsible for the town Party committee and the discipline inspection committee, and the discipline inspection committee reports on the situation of hard work and clean government and accepts quantitative assessment.” Township-equivalent supervision extends to rural communities, with the village discipline inspection committee serving as the township government’s representative at this level [15,16].

After initial results, the reform was gradually rolled out nationwide.

Supervision down to the countryside (SDC) is essentially a reform of the villagers’ self-governance system, and a grassroots policy adjustment to change the prominent problems of villagers’ self-governance. This approach does not refute the validity of villagers’ self-governance, but addresses its current challenges.

Contemporary village power oversight exhibits these features: First, public authority concentrates on collective affairs while strengthening internal–external monitoring, aiming to harness synergistic oversight effects and integrate top-down intra-Party supervision with bottom-up self-supervision for people. Second, the model shifts from static joint monitoring to functional dynamic supervision. External interventions constitute supportive institutional change, requiring constructive engagement among rural societal actors—internal organizations, the two village committees, and villagers—to optimize governance through collaboration and contention. This necessitates the clarification of force responsibilities and the establishment of coordination mechanisms.

2.2. Literature Review

Research on supervisory mechanisms in village governance can be categorized into two distinct types: formal institutional oversight and informal organizational supervision.

Formal institutional oversight examines established regulatory frameworks, specifically focusing on institutional reforms, village financial transparency, electoral procedures, and the consolidated leadership system ‘yijiantiao’ (where a single individual holds both Party and administrative leadership roles) [17,18,19]. Research on informal organizational oversight emphasizes societal actors, including clan participation, elite-led governance structures, transparency in non-profit organizations, and participatory democratic mechanisms among villagers [20,21,22,23,24,25]. Recent research highlights the synergistic effects of formal and informal supervisory mechanisms on rural governance. Clan networks, for example, may subvert formal property rights frameworks by functioning as transactional intermediaries in land redistribution, thereby destabilizing institutional hierarchies [20]. Additionally, formal institutional interactions exhibit geospatial variability and nonlinear developmental pathways [21]. The integrated supervision model analyzed here synthesizes vertical bureaucratic oversight (e.g., superior governmental authorities) and horizontal participatory monitoring (e.g., villager assemblies). Operationally, it institutionalizes village affairs management through committees composed of resident representatives and Party disciplinary officers, standardizing decision-making protocols while bridging the implementation gap in Party-led grassroots governance.

Chinese villages have two main governing bodies that are formally recognized by the state: the village committee (VC) led by the village head (VH) and the village Communist Party committee led by the village Party secretary (VPS). Village Party committees are the organization through which the Chinese Communist Party (CCP) penetrates rural communities. In principle, the village Party committee, and usually the VPS, is to supervise the VH rather than be directly involved in rural governance. In practice, the VPS exerts enormous influence within village governance, thus hindering effective leadership by elected village officials [26]. The co-existence of VCs and Party committees forms two separate loci of power, creating a ‘dual-power structure’ in Chinese villages [27,28]. Empirical studies indicate that the balance of power between the two formal institutions varies across villages: ultimate authority over village affairs may reside with the VPS, the VH, or entail collaborative determinations when the VPS and the VH make joint decisions [29,30].

2.3. Theoretical Analysis

This paper establishes a theoretical analysis framework based on the principal-agent theory and the public governance theory. There exists a diversified grassroots governance organizational system in rural areas of China. The village committee plays multiple roles in rural areas. It is both the head of the village and the agent of the state [31]. The village Party organization connects the higher-level government and the village cadres, and the village-affairs supervision committee builds a bridge for communication between villagers and the village committee. Public governance is a continuous process in which multiple subjects interact and negotiate to solve public affairs for a common goal [32]; thus, multiple subjects need to be integrated into the analytical framework. Supervision down to the countryside, as a reform of the village-level power supervision system, has a significant impact on the supply of public goods in rural areas. The analysis is specifically conducted from three dimensions: supply system environment, supply subject, and supply effect (Figure 3).

The supply environment is an integration of political power and governance power. Supervision down to the countryside (SDC) involves decentralizing supervisory authority from central governments and Party organizations to village administrations, systematically addressing governance bottlenecks in rural public service delivery [33]. The dual-channel design integrates participatory monitoring by villagers (facilitating collaborative resource governance) with hierarchical compliance enforcement through county-level disciplinary inspection systems and township Party committees [34,35].

Well-regulated government oversight helps standardize village officials’ conduct, improves transparency in public affairs and financial management [36,37], and facilitates collective action through community trust [38]. These mechanisms are critical for enhancing public goods provision and economic development in rural areas of developing countries. Therefore, effective village supervision contributes to better management of collective resources and promotes efficient rural governance.

The SDC policy strengthens executives of the VPC and the VSC and increases the participation of households. Village-level power supervision interferes with the behaviors of village-level organizations and village cadres, thus affecting the quality of the village public goods supply. Under the current system, the village committee seems to act as a dual-role agent for the government and villagers. Although it has limited capacity to provide social-security-related public goods independently, it plays a crucial part in the government-led supply process, partially fulfilling the entrusted functions of supplying such public goods [9]. State power in rural communities through various means induces bureaucratic tendencies among village cadres and the semi-bureaucratization of rural governance. This phenomenon explains the involution of village funds, manifested in the rigidification of expenditure structures, rising governance costs, and the encroachment of administrative affairs on daily village activities [7,39].

The implementation of the SDC policy has reinforced higher-level governments’ authority in village governance, imposing constraints on village cadres’ administrative decision-making and financial expenditures. This has resulted in reduced autonomy for local cadres, who exhibit excessive caution, even in routine village spending. The SDC policy also weakens grassroots autonomy by shifting cadres’ priorities toward completing specialized tasks assigned by superiors and avoiding mistakes, often at the expense of community engagement and mobilizing public participation in governance. Complex approval procedures for village expenditures mandated by higher authorities have incentivized cadres to minimize local spending to avoid accountability risks. Accordingly, the following research hypothesis is proposed:

Hypothesis (H1):

The policy of SDC significantly reduces rural public goods expenditure.

The impact of the SDC policy on village public expenditure operates through two pathways: First, the Governance Standardization Effect. The SDC policy standardized grassroots governance and strengthened constraints on village committees and cadres. It reduced cadres’ discretion in decision-making, leading to decreased overall village expenditure. Enhanced procedural requirements limit cadres’ operational autonomy. Under accountability pressures, committees shift from credit-seeking to risk-averse behaviors [40], which discourages necessary expenditure and investment in public goods welfare budgets [41]. Second, the Procedural Complexity Effect. The SDC policy lengthened the approval procedures for public projects, reducing village-level public investments within comparable timeframes. Rural public funding at the time primarily relied on county/township budgets, which faced fiscal constraints due to limited revenue sources. Central and provincial transfers only covered basic governance and education needs, leaving insufficient funds for infrastructure and public services [42]. With tightened budget scrutiny under vertical oversight, township governments approved fewer public projects, further reducing new public investments. Accordingly, this study proposes the following two hypotheses:

Hypotheses 2 (H2):

The Governance Standardization Effect negatively moderated the impact of the SDC policy on rural public goods expenditure, which means overall village expenditure decreased after the policy was implemented.

Hypotheses 3 (H3):

The Procedural Complexity Effect negatively moderated the impact of the SDC policy on rural public goods expenditure, which means the average public investment per village and the number of projects decreased after the policy was implemented.

3. Research Design and Methods

3.1. Data Sources

The data used in the study were from five waves of the China Rural Development Survey (CRDS household data) conducted in 2005, 2008, 2012, 2016, and 2019. (This survey is also known as the “One Hundred Village Survey”; it covers 100 villages as samples). It is a tracking survey; however, it is not a balanced panel dataset due to sample attrition.

The research team members conducted five follow-up field surveys to sample the villages and obtained a set of long-term panel data. The initial survey in 2005 used a stratified random sampling procedure to select a nationally representative sample covering 100 villages in 50 townships from 25 counties across 5 provinces. It took a four-step approach:

First, five sample provinces were selected from each of China’s major agro-ecological zones from a list of provinces arranged in descending order of gross value of industrial output. These were Jiangsu, Sichuan, Shaanxi, Hebei, and Jilin. Second, within each sample province, counties were evenly ranked and divided into five quintiles in descending order according to the per capita gross industrial output (the highest 20% of the per capita gross industrial output values of each county make up the first group, the second 20% of the second-highest industrial output values of each county make up the second group, and so on). One county was randomly selected from each quintile as the sample county. Third, within selected sample counties, each township was divided into two groups in descending order according to the per capita gross industrial output: a “more well-off” group and a “poorer” group. The team chose one from each of the above two groups so that two townships were selected. Fourth, within each township, two villages were selected, following the same procedure. The survey teams selected 20 villages in each province. After processing, 100 villages (1 province × 5 counties × 2 townships × 2 villages) were used for subsequent analysis.

This study uses survey data collected from 2005 to 2019. The specific survey, completed by village officials, gathered information on income and expenditure, per capita income, public utility investment, population, arable land, etc. The questionnaire provides further data regarding village cadres, covering aspects like personal profiles, salaries, evaluation criteria, and incentives and sanctions. Concerning the onset time of the “rural supervision implementation” in each village, the author conducted subsequent telephone inquiries and interviews with village cadres for verification. Additionally, relevant information from county-level official government websites was incorporated for supplementation and cross-verification, aiming to minimize data error.

3.2. Definition of Variables

3.2.1. Dependent Variable

The dependent variable was village public goods expenditure. Based on previous research [43], the study used the “village public service expenditure (infrastructure)” item from the financial section of the village-related questionnaire as a variable to gauge the public goods supply of the village committee. The village’s collective organizations allocate funds for the construction of rural public infrastructure and public welfare undertakings and provide social security for the construction of various people’s livelihood and social undertakings. These public services ensure villagers’ basic living standards and are reflected in the village’s public expenditure.

3.2.2. Independent Variable

The independent variable was the supervision down to the countryside (SDC) policy. The study constructed a dummy variable to quantify the net impact of the policy within a multi-period DID model. This variable indicates whether the “rural supervision policy” has been implemented in the sampled villages. To determine this, several inquiries are utilized:

First, “Treat” is constructed based on two criteria: (1) whether the director of the village supervisory committee concurrently holds the position of village disciplinary inspection committee member; and (2) whether the director is subject to supervisory oversight by superior government entities (including township/county discipline inspection commissions and supervision offices). “Treat” assumes a value of 1 if both conditions are simultaneously satisfied within a village, and 0 otherwise. Second, the following inquiry is utilized: “Has this reform taken place in the village during the current year?” (denoted by “post”).

If the answers to “Treat” and “post” are affirmative (coded as 1), the village is classified as having implemented the SDC policy and thus belongs to the experimental group. Conversely, if either answer is negative (equals to 0), the village is considered not to have fully implemented the policy and is assigned to the control group.

3.2.3. Mechanism Variable

The mechanism variable was overall village expenditure and village public investment projects. Referring to the questionnaire, village expenditure is selected as an indicator to measure the extent of budget constraints. In the questionnaire, the two indicators of “new public investment in villages” and “number of beneficiaries of public projects” can also reflect the use of funds for public goods in villages.

3.2.4. Control Variables

Self-sufficiency of village public goods is often related to the natural characteristics, economic level, and development of the village. Moreover, characteristics of village cadres, external policies, and other factors may also influence the self-sufficiency of the public goods supply. Based on the paper of Li (2016), to select control variables and incorporate the discussion of factors influencing rural economic development by Zeng (2012) [44,45], we chose relevant control variables from four dimensions: natural geographical conditions, socioeconomic attributes, village internal governance, and institutional arrangements.

For natural geographical conditions, two variables were selected: plowland area and the distance between the village committee and the township government. Socioeconomic attributes mainly included villagers’ per capita net income, the number of enterprises in the village, total population, and “yijiantiao”. For village internal governance, we chose the characteristics of village cadres, like village cadres’ education level, whether they worked outside or not, whether they worked within a self-employed industry/commerce or not, and whether they belonged to the surname of the village or not. Institutional arrangements reflect the support of higher government, and this paper chooses “transincome” as a variable.

Table 1. Descriptive statistics of the variables.

Variable	Code	Obs	Mean	S.D.
Dependent variable
Village public service expenditure	publicexp1	488	7.404	26.794
Independent variable
Supervision down to the countryside	DID	495	0.162	0.368
Natural geographical conditions
Plowland area	plowland	490	2722.673	2728.128
Log (plowland area)	lnplowland	490	7.353	1.404
Distance between the village committee and the township government	distance	499	5.786	6.859
Socioeconomic attributes
Villagers per capita net income	fincome	494	6594.478	5623.962
Log (villagers per capita net income)	lnfincome	494	8.463	0.849
the number of enterprises	enterprise	490	3.943	9.363
“Yijiantiao” (whether the Party secretary and the village branch secretary are the same or not)	election	491	0.334	0.472
Total population	population	494	1587.221	1070.239
Crackdown activities	crackdown	375	0.387	0.488
Number of reactionary organizations	reactionary	375	1.851	3.540
Villager disputes	villager disputes	370	0.378	0.486
Clan conflicts	clan conflicts	365	0.186	0.390
Village internal governance: characteristics of village cadres
Belongs to the surname of the village or not	daxing	493	0.416	0.493
Education	edu	493	3.560	0.816
Working outside or not	outside	493	0.183	0.387
Working in a self-employed industry or not	business	492	0.285	0.452
Institutional arrangements
Transfer payment	transincome	495	10.161	31.609
Village collective economic income	CEI	486	30.215	82.959
Log (village collective economic income)	lnCEI	486	2.290	1.389
Mechanism variable
Total expenditure of the village	Exp	496	26.196	68.064
New investment in economic infrastructure	NIE	474	39.124	83.953
New investment in social-security projects	NIS	488	215.214	220.239
Number of people benefited by economic infrastructure	NPE	496	31.5427	373.6279
Number of people benefited by social-security projects	NPS	483	72.504	180.730

Notes: ① The data presented above are derived from the survey samples (CRDS) through relevant calculations and arrangements. ② Logarithmic values of the village cultivated-land area and per capita net income are utilized for regression analysis. ③ Since 2013, owing to adjustments in national statistical indicators, the annual per capita disposable income has been adopted.

reports the descriptive statistics for all the variables.

3.3. Methods and Model Specification

Using the multi-time point differential identification strategy, this study empirically investigates whether the SDC policy significantly affects rural economic development to test the hypothesis. The two-way fixed effects model set used is as follows:

$${publicexp}_{it}={\alpha }_0+{\beta }_1{DID}_{it}+{\beta }_2{X}_{it}+{\phi }_i+{year}_t+{\epsilon }_{it},$$

(1)

$$M_{it}={\alpha }_2+{\beta }_3{DID}_{it}+{\beta }_4{X}_{it}+{\phi }_i+{year}_t+{\epsilon }_{it},$$

(2)

where ${publicexp}_{it}$ represents the public expenditure of sample villages from villages in the year (2005/2008/2012/2016/2019); and ${DID}_{it}$ is a dummy variable, indicating whether the village from sample i has implemented the supervised rural policy. If yes, the value is equal to 1, which is also the main explanatory variable concerned in this paper. Coefficient ${\beta }_1$ reflects the net effect of the implementation of the policy. $M_{it}$ represents the mechanism variable, which includes the total expenditure level of the village and new public investment projects in the village; and $X_{it}$ is a series of control variables. ${\phi }_i$ and ${year}_t$ control the village-level fixed effect and annual fixed effect, respectively, so as to avoid the result bias caused by the possible systematic differences in economic level, administrative level, and natural and social capital accumulation of villages between the villages that have implemented the supervision and the villages that have never implemented the policy. ${\mathsf{\epsilon }}_{\mathrm{i}\mathrm{t}}$ is the random error.

4. Empirical Analysis

4.1. Baseline Regression

This study introduces control variables from four dimensions (natural geographical conditions, socioeconomic attributes, village internal governance, and institutional arrangements) to analyze the impact of supervision down to the countryside on the public expenditure of villages.

Table 2. The results of the baseline regression.

Variable	(1)	(2)
DID	−5.857 **	−6.712 *
	(2.558)	(3.519)
lnplowland		−2.117
		(1.572)
distance		−0.253 ***
		(0.076)
lnfincome		−2.548
		(2.297)
enterprise		−0.117
		(0.086)
“yijiantiao”		6.921
		(4.656)
total population		0.012 *
		(0.007)
daxing		0.001
		(2.006)
edu		1.763
		(2.156)
outside		−1.189
		(1.969)
business		−2.588
		(2.930)
transincome		0.289
		(0.268)
Constant	1.549	13.880
	(1.294)	(18.800)
Village fixed effect	Yes	Yes
Year fixed effect	Yes	Yes
R2	0.123	0.338
Obs	489	469

Notes: ① Regarding the “distance between the village committee and the township government” variable in the regression table, it was repeatedly surveyed. The relocation of certain township governments led to numerical changes in this variable. Consequently, even after controlling for village and year fixed effects, the coefficient may remain significant. ② Robust standard errors, clustered at the village level, are presented in parentheses. *** p < 0.01, ** p < 0.05, * p < 0.1.

reports the regression results for the supervision policy. The results show that with the gradual inclusion of control variables, the estimated coefficients of the supervision variable remain significant and negative, indicating that collective action can boost rural economic development, supporting H1.

As indicated by the results presented in columns (1) and (2), upon incorporating village-level and year fixed effects, the regression analysis reveals that the supervision policy exerts a notable negative influence on village public expenditure. Specifically, the regression coefficient for the supervision variable is −5.857, significant at the 5% level. After controlling for variables, the absolute value of the policy effect increases, with the regression coefficient reaching −6.712 and remaining significant at the 10% level. This implies that supervision policy bolsters both intra-Party and democratic oversight of village cadres. In villages where this policy is implemented, the supply of public goods is significantly curtailed. On average, expenditure for maintaining and managing public infrastructure declines by 67,120 yuan.

Based on the regression outcomes of the control variables, an increase in the distance between the village committee and the township government leads to a significant reduction in village public goods expenditure. This indicates that the greater the separation between the village committee and the township government, the less constrained the village committee is by the township government. As a result, village cadres have more leeway to allocate funds to other projects at their discretion.

The SDC policy has generated a constraining effect on village cadres. By strengthening oversight over village cadres, SDC triggers cautious rational responses among them. Cadres adopt more conservative strategies, complying with supervision requirements in the short term. For instance, they curb superficial formalities at the grassroots level and reduce non-essential expenditures. They also rigorously control non-compliant projects and strictly manage income and expenditure records, thereby enhancing village financial transparency and village affairs management capacity.

4.2. Parallel-Trend Test

To validate the efficacy of the identification strategy, the selected samples need to satisfy the parallel-trend assumption. The study implies that prior to the policy shock of the implementation of the supervision policy, there was no significant difference in the development trends between the treatment group and the control group, so as to ensure the accuracy of the DID estimation results. Hence, the paper conducts a parallel-trend test on the aforesaid outcomes.

The initiation times of the rural supervision policy pilot vary across villages. In this section, dummy variables spanning from the third (−3) period prior to the policy implementation to the third (+3) period after it are incorporated into the original model. The results of the parallel-trend test are presented in Figure 4. The policy time points represent the time lags in the implementation of rural supervision among different sample villages. For instance, the 0 period indicates that the sample village implemented rural supervision during the survey period; the (−1) period corresponds to the survey immediately preceding the village’s adoption of rural supervision; and the (+1) period represents the first survey after the village’s implementation of rural supervision.

As depicted in Figure 4, the estimated coefficients for the periods from the (−3) period to the (−3) period before the policy time point are not significantly different from zero within the 90% confidence interval. This suggests that no significant differences existed between the treatment and control groups prior to the implementation of rural supervision, thereby validating the parallel-trend assumption. Additionally, starting from the 0 period (i.e., upon policy implementation), the estimated coefficients are significantly below zero, indicating a substantial negative policy effect.

4.3. Placebo Test

To protect against the possible impact of omitted variables on our results, we conduct a placebo test by randomly choosing the village as the treated village.

A placebo test is conducted by randomly sampling a treatment group from the dataset. The sample encompasses 100 villages, with 56 having implemented the supervision policy. For the placebo test, 56 villages are randomly selected from the 100-village sample to form a pseudo-treatment group. Given the application of the multi-period DID method, which implies multiple implementation time points for the supervision policy, each village in the pseudo-treatment group is randomly assigned a policy-implementation time point. This process is replicated 500 times. Subsequently, the t-statistics of the post-policy period from these 500 regressions are compared with those of the benchmark regression.

As illustrated in Figure 5, among the 500 regression results, only a negligible number of cases exhibit a t-statistic for the supervision policy that exceeds the post-implementation t-statistic of the benchmark regression (red line value: −1.91). The red line represents the t-statistic values of the coefficients of the independent variables in column 2 of the benchmark regression (Table 2).This indicates a low probability of committing a Type II error. Thus, it can be inferred that the negative impact of the supervision policy on the village committee’s involvement in public goods supply is robust, effectively reducing the village’s investment in public projects.

4.4. Robustness Test

This section will discuss the robustness of empirical results through the exclusion of sample selection bias and other policy effects resulting in estimation bias.

First, the data in this paper span from 2005 to 2019, covering a relatively long time period. During this period, villages have also experienced many changes. Thus, the sample estimation results will be affected by many other policies. The paper adds the interaction fixed effects of counties and years for regression to control the impact of other policies on the results. The results in column 1 of

Table 3. Robustness tests.

Variable	(1)	(2)	(3)	(4)
DID	−11.584 *	−11.337 *	−14.733 *	−8.185 *
	(6.384)	(6.305)	(7.537)	(4.562)
Controls	Yes	Yes	Yes	Yes
Village FE	Yes	Yes	Yes	Yes
Year FE	Yes	Yes	Yes	Yes
County × Year FE	Yes	Yes	Yes	Yes
Obs	469	469	450	469
R2	0.388	0.389	0.394	0.378

Note: Standard errors are in parentheses. * p < 0.1.

represent that the regression coefficient of the SDC policy reaching the village is significant, negative, and its absolute value increases.

Second, some of the villages in this paper are poverty-stricken villages, and they successively lifted themselves out of poverty before 2019. Therefore, there may be an estimation bias caused by the different development priorities of the SDC policy reaching the village before and after the villages emerged from poverty. In addition to controlling the fixed effects of “county × year”, this paper adds the control variable of “whether the village is a poverty-stricken village” and then conducts the regression. Regression results are shown in column 2 of Table 3. The absolute value of the estimated coefficient of supervision reaching the countryside has decreased, but it is still significantly negative.

Third, the earliest time for a region to start the pilot was in 2010, according to the official announcements of governments. However, four villages had already started to implement this reform in 2008. This may be caused by different survey inquiry scopes. To ensure the reliability of the results, this paper excludes the villages that implemented the policy in 2008 from the samples and then conducts the regression. It can be seen from the results that the impact of the implementation of the SDC policy on the village public goods expenditure is significantly negative, verifying the robustness of the research conclusions in column 3 of Table 3.

Fourth, regarding the rural consumer price index (CCPI): the village expenditure data obtained from the survey are all nominal amounts, not the real amount. Thus, the paper selects the “rural consumer price index (CCPI)” for conversion to obtain the actual expenditure amount, and then substitutes it into the model for regression. The results are shown in column 4 of Table 3. After the implementation of the policy, the public expenditure of villages on welfare security and infrastructure maintenance decreases by nearly 80,000 yuan, which is consistent with the benchmark regression, and the coefficients are close.

Fifth, there is a sample self-selection issue. Some villages may be more likely to obtain ‘supervision-to-village’ pilot status due to their socioeconomic characteristics. To exclude this influence, this paper conducts the following analyses:

First, positive selection bias. Villages with better socioeconomic conditions could actively seek policy inclusion, making affluent villages more likely to obtain “supervision-to-village” pilot status. To test this, we regress the dependent variable “whether a village implemented SDC” on “village total income” using Model (1). Insignificant coefficient would indicate no correlation between SDC selection and village economic status. Second, negative selection bias might occur. High-conflict villages could be prioritized for supervision reform. We measure “historical village conflict” as the independent variable (proxied by four indicators: crackdown activities, number of reactionary organizations, villager disputes, and clan conflicts). Regressing this on “SDC implementation status” in Model (1), an insignificant coefficient would suggest no link between SDC selection and village conflict levels. As shown in

Table 4. Robustness tests: self-selection test.

Variable	DID (Whether a Village Implemented SDC)
	(1)	(2)	(3)	(4)
crackdown	0.00762
	(0.00919)
reactionary		0.00328
		(0.00387)
villager disputes			0.00680
			(0.00864)
clan conflicts				−0.00239
				(0.0130)
lnCEI	0.0149	0.0150	0.0135	0.0151
	(0.0127)	(0.0127)	(0.0131)	(0.0133)
Controls	Yes	Yes	Yes	Yes
County FE	Yes	Yes	Yes	Yes
Year FE	Yes	Yes	Yes	Yes
County × Year FE	Yes	Yes	Yes	Yes
Obs	344	344	339	334
R2	0.7712	0.7713	0.7697	0.7709

Note: Standard errors are in parentheses.

, coefficients for both village income and historical conflict are statistically insignificant. This implies that SDC villages were not selected based on affluence or conflict intensity, ruling out positive and negative selection biases. Third, dynamic selection bias. After SDC policy implementation, initial successful villages might trigger imitation by neighboring villages. However, this study’s sample exclusively comprises non-adjacent villages, thereby addressing this bias at the research design stage.

To mitigate the potential impact of sample self-selection on the regression results, this paper further employs a propensity score matching difference-in-differences (PSM-DID) approach to examine the effects of the SDC policy.

Three matching methods were used in the analysis, and the matching results are presented in

Table 5. Matching quality test: balancing property.

Sample	Pseudo R2	LR chi2	p > chi2	MeanBias	MedBias
Before matching	0.199	78.60	0.000	27.5	17.0
Nearest neighbor matching (k = 2)	0.011	2.07	0.990	4.9	3.7
Kernel matching (bw = 0.03)	0.004	0.83	1.000	5.5	5.7
Radius matching (bw = 0.02)	0.006	1.19	0.999	5.5	4.9

. Propensity score matching was applied to address potential biases by ensuring comparability between the treatment and control groups. The common support condition was used as the primary test to evaluate the effectiveness of the matching procedure. As shown in Table 5, the matching process resulted in lower Pseudo-R², mean bias, and median bias compared to the pre-matching values, indicating a better balance between the treatment and control groups.

Relevant test results are displayed in Figure 6 (left: standardized differences; right: common trends). Figure 6 (Panels ACE) shows that before matching, most variables exhibited large standardized differences, whereas after matching, the standardized differences for all variables were substantially reduced and fell below 10%, meeting the requirements for a randomized experiment. Furthermore, Figure 6 (Panels BDF) indicates that most observations lie within the common support range, implying minimal sample loss during matching.

The regression results from the PSM-DID estimation are presented in

Table 6. Estimation results of PSM-DID.

Matching Method	ATT Effect	Std Error	T-Value	Controls	Treated
PSM (Nearest neighbor matching)	−11.9427 **	5.5479	−2.15	366	71
PSM (Kernel matching)	−11.4947 **	5.27498	−2.18	384	72
PSM (Radius matching)	−12.0210 **	5.3611	−2.24	384	71

Note: ** p < 0.05.

. The results show that the core explanatory variable remains statistically significant and negative across all matching methods.

4.5. Heterogeneity Analysis

This study examines the heterogeneity impacts of the supervision policy on different sample groups from three dimensions: economic development (this study takes “whether the region was once a national poor county” as a measure to distinguish the level of economic development. If it was, the level of economic development is weak; otherwise, it is stronger), factional conflicts, and the evaluation of village cadres. The regression results are presented in

Table 7. Heterogeneity analysis.

Variable	Classification
	Economic Development		Factional Conflicts		Evaluation of Village Cadres
	Weak	Stronger	Yes	No	Public Goods	Others
	(1)	(2)	(3)	(4)	(5)	(6)
DID	2.325	−19.752 **	1.563	−17.205 **	7.338 ***	−15.77 *
	(4.170)	(9.286)	(5.733)	(8.371)	(2.101)	(8.903)
Control Var.	Yes	Yes	Yes	Yes	Yes	Yes
Village FE	Yes	Yes	Yes	Yes	Yes	Yes
Year FE	Yes	Yes	Yes	Yes	Yes	Yes
County × Year FE	Yes	Yes	Yes	Yes	Yes	Yes
Obs	113	357	134	336	114	356
R2	0.466	0.428	0.450	0.457	0.913	0.453

Note: Standard errors are in parentheses. *** p < 0.01, ** p < 0.05, * p < 0.1.

.

As shown in columns 1 and 2, in regions with weak economic development, this is due to lack of capital, resources, and talent. After the implementation of the SDC policy, the use of funds by village cadres is greatly restricted, and coupled with the lack of professional guidance, the members of the village committee may be more willing to use the money for other purposes such as economic development and administrative expenses, and the village’s money for public expenditure is thereby significantly reduced.

As shown in columns 3 and 4, the level of public expenditure decreased significantly in villages that had not experienced factional fighting, while those that had experienced factional fighting were unaffected. This shows that the supervision policy has a more obvious negative impact on rural public goods supply for harmonious villagers.

The evaluation standard of village cadres shows obvious differences related to the effect of the supervision policy. Where public service constitutes the primary performance metric for village cadres, the influence of the SDC policy on the village’s public service expenditure is significantly increased (column 5 of Table 7). In other villages, the impact of supervision on local public service expenditure remains significantly negative, consistent with baseline regression (column 6 of Table 5). This suggests that the actual effect of surveillance policies is also constrained by other assessments.

5. Mechanism Analysis

5.1. Regression Analysis: Budget Constraint Mechanism

The study incorporates variables like total village expenditure, new investment in economic infrastructure, new investment in social-security projects, number of people benefited by economic infrastructure, and number of people benefited by social-security projects as mechanism variables. The study aims to delve into the internal transmission mechanism through which the supervision policy influences village-level public expenditure.

The policy of SDC has created a constraining effect on village cadres’ official duties. The SDC policy has strengthened the supervision over village cadres in villages, stimulating their defensive and rational responses, and they have adopted more conservative strategies.

Column 1 of

Table 8. Mechanism analysis.

Variable	Exp	NIE	NPE	NIS	NPS
	(1)	(2)	(3)	(4)	(5)
DID	−34.930 *	−44.519 *	−59.646	73.272	−64.994 *
	(17.671)	(26.131)	(50.452)	(52.911)	(35.102)
Control Var.	Yes	Yes	Yes	Yes	Yes
Village FE	Yes	Yes	Yes	Yes	Yes
Year FE	Yes	Yes	Yes	Yes	Yes
County × Year FE	Yes	Yes	Yes	Yes	Yes
Obs	477	0.279	0.310	0.141	0.175
R2	0.497	457	469	477	464

Note: standard errors are in parentheses. * p < 0.1.

shows that the SDC policy significantly reduced total village-level expenditure by 349,300 yuan, indicating substantial constraints on village cadres’ official conduct. This result aligns with the defensive strategy theory under the principal-agent framework [46]. By intensifying oversight (e.g., full financial disclosure and strict project approval), the policy increased the traceability of cadres’ actions, prompting risk-averse strategies [47]. Amid rising uncertainty, cadres prioritized cutting ambiguous expenditure (e.g., non-essential public services) to mitigate accountability risks, ultimately contracting total spending. Thus, increased accountability risks triggered defensive contraction behavior by village cadres. As a result, the total village expenditure declined, thus verifying H2.

Supervision in rural areas promotes the standardization and normalization of village public affairs procedures, but also leads to slower approval speeds for village public projects. Investment amounts in new economic-infrastructure public goods, new social-service public goods in the village, and the number of beneficiaries of these projects, are introduced as variables representing the village public goods investment level. The test results are presented in columns 2 to 5 of Table 8.

Columns 2 and 3 of Table 8 reveal that the policy significantly reduced investments in economic infrastructure, particularly in new roads and power facilities. It reflects the “procedural legitimacy paradox” in organizational theory [48]. While mandating full compliance in project approvals (multi-level co-signatures and villager voting) enhanced procedural rigor, complex processes prolonged decision-making chains. To avoid procedural noncompliance risks, cadres favored low-decision-complexity projects (maintaining existing facilities like road repairs) over new constructions [49]. Thus, procedural formalization incurred opportunity costs for infrastructure renewal, despite curbing corruption.

Columns 4–5 of Table 8 demonstrate that the policy significantly decreased the number of social welfare beneficiaries, while nominal investment remained unchanged. This supports the administrative burden theory [50]. Strengthened supervision required multi-tiered verification (democratic appraisals and public disclosure), creating psychological and compliance costs—termed “eligibility delays” [51,52]. Villagers abandoned applications due to information asymmetry or procedural costs, leading to unspent funds and reduced coverage, despite stable budgets.

The lag effect in the use of project funds led to a relative decline in the number of farmers receiving immediate services and support, verifying H3.

5.2. Case Analysis: Further Discussion on the Implementation of Constraint Mechanisms

BJ Village is situated at the northernmost tip of Sanjie Town, Shaoxing City, Zhejiang Province. It has a population of 2498 from 808 households, with 1626 mu of cultivated land, of which 1323 mu are paddy fields and 300 mu are dry land, resulting in a cultivated land per capita value of less than one mu. The village’s collective economic income exceeds one million yuan. BJ Village has been successively named a demonstration village for rural-level Party-style and clean-governance construction in Zhejiang Province. In 2010, a village-affairs supervision committee was established, establishing an autonomous organizational system. In this system, the village Party organization serves as the leading core, the villagers’ meeting and villagers’ representative meeting act as the main decision-making and deliberative bodies, the villagers’ committee functions as the management and execution body, and the village-affairs supervision committee takes on the role of the supervision body.

However, due to the meticulous and cumbersome nature of village-affairs work, village cadres sometimes actively seek personal benefits. The manifestations are as follows:

First, irregular practices in village cadres’ management of projects. For projects with a total budget below a certain threshold, the village can self-organize small projects. Sometimes, construction projects do not follow the project-based operation procedures strictly, and they are executed first and then inspected, which provides opportunities for village cadres to gain hidden benefits. Second, there are benefit-seeking issues in the implementation of bidding projects. Most construction companies rely on village cadres’ assistance to carry out projects, so village cadres may claim personal benefits. Third, the projects lack supervision and thus the financial situation is unclear. Expenditures for village projects and daily operations are not well-regulated, and the problem of non-standard spending is hard to solve fundamentally.

To strengthen the supervision of grassroots public power, the region has reformed the grassroots discipline inspection and supervision system, and BJ Village has followed suit. After the implementation of rural supervision, the village restricts village-level power through institutional norms and mass supervision, standardizing the job-related behaviors of village cadres. The details are as follows:

First, the operation process is standardized. Public projects must be publicly tendered. For invoice reimbursement over 2000 yuan for project implementation and management, the signatures of five people, including the village secretary, the sign-approver, the transaction handler, the deputy village head, and the director of the village-affairs supervision committee, are required. Small-scale projects independently arranged by village organizations must be voted on by a show of hands at the villagers’ representative meeting and can be implemented only after passing the publicity period. All financial expenditure must be publicized, prolonging the processing cycle of some village affairs. The village has established a comprehensive system, including integrity files, for public officials, and an information database for supervision objects, and has strictly enforced the supervision and management of village-level “three resources”, minor powers, and project bidding, ensuring full coverage of the basic information of supervision objects and real-time dynamic updates of various data. Thus, the probability of non-standard expenditures being approved has been significantly reduced (the downward arrow in Figure 7 indicates a decrease in expenditures).

Second, the SDC policy is actively mobilizing villagers’ participation. The village has formed a team of supervision informants and developed a duty roster for them, clarifying work responsibilities, duty execution methods, and assessment, reward, and punishment mechanisms. The team consists of grassroots village-stationed cadres, the director of the village supervision committee, and retired cadres. Members of the village supervision committee are selected from prominent families, local sages, and villagers with a history of petitioning in the village. Members of the village discipline inspection committee can also serve as the director of the village supervision committee. On the one hand, the supervision by the village-affairs supervision committee covers entrepreneurial commitments, construction projects, convenience services, and village-level finances. On the other hand, committee members need to closely interact with the masses to understand their actual demands.

Here is the analysis diagram of the conduction mechanism (Figure 7).

Drawing on the “state suspension–embeddedness” theory in political economy, this study examines the inherent tensions among oversight mechanisms, community autonomy, and political legitimacy through the case of BJ Village (

Table 9. Analysis on the “state suspension–embeddedness” theory.

Mechanism	Measures	Tension Manifestation	Theoretical Lens
Procedural formalization (institutional embeddedness)	-Reimbursement approval requires the signatures of five people -Small projects require villager assembly approval -Full financial disclosure	Autonomy erosion: -Administrative burden ↑ -Public service time ↑	Rules penetrating local practice
Participatory oversight (participatory embeddedness)	-Supervisors include petitioners/village elites -Discipline committee director chairs oversight committee	Legitimacy trade-off: -Cadre accountability ↑ -Elite capture risk	State-engineered legitimacy building

Note: The arrow (↑) means increase.

).

The SDC policy examined here embodies this duality through two institutional innovations: procedural formalization (institutional embeddedness) and participatory oversight (participatory embeddedness).

As a provincial model of rural governance, BJ Village established a self-governance framework by 2010 (village Party, village committee, village supervision committee), yet suffered from suspended governance: cadres profited from “construct-first-bid-later” loopholes in small projects, extracted bribes from external contractors, and perpetuated irregular expenditures due to oversight deficits. Post-2010 reforms deployed dual embeddedness: institutional embeddedness mandated procedural rigidity (e.g., reimbursement approval requires the signatures of five people, villager assembly approval for small projects, financial disclosure), while participatory embeddedness created oversight teams co-opting petitioners/village elites and merged disciplinary roles into oversight committees to implement “mass-controlled funds”.

This study integrates data from interviews with village cadres and surveys of farming households to further analyze their respective perspectives on the supervision down to the countryside (SDC) policy.

The following is an excerpt from an interview with the Party secretary of BJ Village:

Both the positive and negative aspects are quite evident. On the positive side, village officials are less likely to make fundamental mistakes, which increases public trust and makes it easier to implement policies and affairs. The supervision of the entire process has become stricter, with more procedures in place. Major matters require approval by the village representative assembly, while minor ones need several days of review. All public projects must go through a series of steps, including public bidding, announcement, filing, and construction, which slows down the process and makes village officials feel constrained. Overall, while the security and social atmosphere will gradually improve, in the short term, this has little significant impact on the village’s economic development.

Based on the interview, the village cadre holds a dualistic perspective on the top-down supervision embodied by the village supervision committee (VSC). The cadre acknowledges its intended benefits, namely enhanced official accountability, reduced fundamental mistakes, increased public trust, and smoother policy implementation. However, he equally emphasizes significant operational drawbacks, primarily procedural burden. This formalization, requiring multi-level approvals and strict adherence to processes like public bidding for all projects, drastically slows down decision-making and implementation, leading village officials to feel constrained in their actions. The response reveals no evidence of overt resistance but rather a clear dynamic of frustrated adaptation. The cadre’s focus on meticulous procedural compliance—navigating the complex rules to avoid individual fault—suggests an informal adaptation, where following the process correctly becomes a primary goal, potentially at the expense of efficiency.

As shown in

Table 10. Cadres’ official conduct and households’ satisfaction.

Variables	Treat = 0	Treat = 1	Difference	T-Value	Data Year
Cadres’ official conduct
Collective asset liquidation	0.9545	1.000	0.0454	1.6200	2019
Households’ satisfaction with public services
Roads	0.5767	0.7117	0.1350 ***	10.0130	2005, 2008, 2012, 2015, 2019
Schools	0.5502	0.5660	0.0158	0.3933	2005, 2008, 2012, 2015, 2019
Clinics	0.6248	0.6982	0.0734 ***	5.4891	2005, 2008, 2012, 2015, 2019
Drinking water	0.7234	0.8532	0.1298 ***	10.4001	2005, 2008, 2012, 2015, 2019
Irrigation systems	0.4731	0.6503	0.1772 ***	12.6497	2005, 2008, 2012, 2015, 2019
Drainage systems	0.4853	0.6299	0.1442 ***	9.8287	2005, 2008, 2012, 2015, 2019

Note: Data were compiled by the author. *** p < 0.01.

, the implementation of the SDC policy has brought changes both in the conduct of village cadres’ official duties and in villagers’ satisfaction with public goods. After the implementation of the policy, village cadres generally performed their duties in accordance with regulations—though the change was not statistically significant—while villagers’ satisfaction with public services increased markedly. Farming households reported significantly higher satisfaction with roads, clinics, drinking water, as well as irrigation and drainage facilities. This improvement is associated with increased trust resulting from cadres’ compliance with regulations under enhanced supervision [53], reflecting, to some extent, farmers’ approval of the policy and a substantial boost in their satisfaction with village governance.

In summary, following the implementation of the SDC policy, BJ Village has regulated the behavior of both village cadres and villagers in two ways: First, it standardized village affairs procedures (procedural formalization). Although this has increased the administrative burden to some extent and slowed down decision-making and implementation, it has also reduced major errors, enhanced public trust, and facilitated policy enforcement. However, village cadres should remain alert to the potential trade-offs in terms of efficiency and local autonomy. Second, it promoted public participation in supervision (participatory oversight). Involving villagers in monitoring not only expands cadre accountability but also improves their satisfaction with village governance due to increased trust, which in turn reinforces the implementation of supervisory policies, creating a positive feedback mechanism.

6. Conclusions and Implications

6.1. Conclusions

Using data on rural public service expenditures and investments from 100 villages in five provinces spanning 2005 to 2019, this study employs a multi-period DID fixed-effects model to explore the impact and underlying mechanisms of the supervision policy (SDC) on the public goods supply of village committees. The findings are as follows:

Firstly, the supervision policy significantly weakens the village committee’s public goods supply. This is evident in the notable decline in village-level public spending. Secondly, the policy’s impact exhibits heterogeneity based on village characteristics. Specifically, the weakening effect is more pronounced in villages with harmonious social relations, in economically underdeveloped areas, and in villages where public-service provision is not the key performance indicator for village cadres. Conversely, in villages where public-service construction is the primary assessment criterion for village cadres, the policy has a positive, promoting effect. Thirdly, adoption of risk-avoidance strategies by village cadres drives the reduction in village public goods supply under the SDC policy. Cadres favored low-decision-complexity projects in order to avoid procedural noncompliance risks, and procedural complexity results in “eligibility delays,” reducing the actual coverage of program benefits. Case analysis indicates that two means of achieving the budget-constraint mechanism are rectifying non-standard public spending through procedural formalization and participatory oversight.

6.2. Implications

The findings of the study bear multiple implications.

Firstly, the study offers crucial practical grounds for the government’s future selective implementation of the rural supervision pilot program. Our results indicate that enhanced supervision of village-level organizations by higher-level governments may dampen village cadres’ work enthusiasm and reduce their engagement in public goods provision. Thus, it is essential to conduct research on issues emerging from the current rural supervision scenario, adhering to the principle of local-condition adaptation rather than blind promotion.

Secondly, an appropriate adjustment of village-cadre assessment criteria is advisable. When confronted with the short-term negative impacts of rural supervision on village-level public spending, this can be counterbalanced by adjusting the assessment criteria, such as by reducing the economic-related evaluation system during the implementation of supervision in rural areas, and increasing indicators related to public services and other aspects of people’s livelihood. Moreover, it is useful to regulate any boundary-overstepping behaviors of village cadres through internal and external assessment and supervision mechanisms.

6.3. Discussion

The study also implies that rural supervision policies need to pay attention to boundaries. It is not the case that the stronger the supervision force, the better the supervision effect. Rural expenditures will be affected; in particular, rural public expenditure will significantly decrease. This provides empirical evidence for the viewpoint in the theory of power supervision regarding “the coupling effect between formal and informal institutions in rural society”. This means that the sustainable development of rural governance not only relies on government policies but also needs to attach importance to the inherent strength of the village society. It contrasts with Jürgen Habermas’ governance theory perspective that “procedural rationality can resolve value conflicts in pluralistic societies,” supplementing the different possibilities of “procedural rationality” in public domain governance. The research found that villages with harmonious relations among villagers can weaken the negative effects of SDC policies, providing empirical evidence for Jurgen Habermas’ governance theory that “rational communication is emphasized as the basis for resolving differences”.

While this study focuses on the micro-level dynamics of village governance and public goods provision under the SDC policy, its findings resonate with broader macro processes of urban–rural integration in contemporary China. The SDC, as a mechanism of bureaucratic recentralization and standardization, can be understood as part of the State’s toolkit to manage the profound challenges posed by rapid urbanization and the uneven geographical distribution of population and resources, famously conceptualized by the Heihe–Tengchong Line [54]. As national policies actively promote population mobility and urban agglomeration while simultaneously aiming to revitalize rural areas, the governance of villages—particularly those in peripheral or economically lagging regions—becomes critically important. The risk-averse behavior of village cadres and the consequent retraction in local public spending, as identified in our micro-analysis, reflect a localized response to these macro national imperatives. The SDC’s procedural formalization, while aiming to improve governance, may inadvertently constrain the adaptive capacity of village institutions precisely at a time when they need flexibility to respond to shifting community needs. Thus, the micro-mechanisms uncovered here are not isolated; they are a facet of the larger state project to recalibrate governance across a rapidly transforming national territory, ensuring stability and minimum service standards.

Our findings suggest that the SDC operates not merely as a tool for improving rural governance but also as a potent instrument of functional recentralization. By imposing standardized procedures and enhancing upper-level government oversight, the SDC strengthens the vertical integration of the State apparatus, potentially at the expense of local autonomy and grassroots initiatives. This trend aligns with a broader shift towards the recentralization of authority in various policy domains [55]. The macro risk lies in the potential homogenization of governance models across vastly diverse local contexts, which could stifle the very innovation and contextual adaptation necessary for effective development. Therefore, a critical academic and policy consideration is how to balance the undeniable benefits of enhanced accountability and anti-corruption offered by such supervisory mechanisms with the need to preserve functional discretion at the local level. Future research should explore how the SDC’s framework interacts with other macro-policies, such as those promoting rural–urban integration, to shape the long-term trajectory of state–society relations and spatial development in China.