The power of the visible hand's micro-intervention in agricultural credit: a policy analysis based on Sichuan Province's risk compensation fund system

This study investigates the effects of the Risk Compensation Fund System (RCFS), a government micro-intervention in agricultural credit, in Sichuan Province, China. Using the staggered difference-in-differences method as a benchmark model, we find that the RCFS increased the added value of the primary industry by 14%. Robustness tests using the Callaway and Sant'Anna (2021) Difference-in-Differences and Bacon decomposition confirm the validity of our hypothesis, suggesting that the RCFS can effectively promote agricultural development. Furthermore, we employ the Synthetic Difference-in-Differences method to estimate the policy's individual treatment effect. Heterogeneity analysis indicates that the RCFS can be successfully implemented even in areas with poor geographical conditions and low levels of economic development, highlighting its potential for policy expansion. However, our findings suggest that policies focusing solely on promoting competitive county-level banking or increasing agricultural leverage do not necessarily lead to optimal outcomes. Instead, policymakers should prioritize improving the efficiency of fiscal fund utilization and enhancing the policy implementation mechanism to optimize the benefits of the RCFS. Our study underscores the positive significance of targeted government intervention in promoting economic development in developing countries and emphasizes the importance of well-designed micro-mechanisms in ensuring policy success.

1 Introduction

The role of the visible hand in economic development has been a subject of widespread debate (Krueger, 1990; Goldman and Gorton, 2000; Acemoglu and Robinson, 2013; Keech and Munger, 2015). However, to achieve industrial development goals, developing countries find it difficult to reject the potential economic development benefits brought by the visible hand. In recent years, China has been committed to enhancing agricultural competitiveness and promoting agricultural transformation through the visible hand, encouraging local governments to innovate policies in agricultural credit investment.

Against this backdrop, China's Sichuan Province has actively pursued policy and institutional mechanism innovations. On November 22, 2018, the Sichuan Provincial Department of Finance, the Department of Agriculture, and the Chengdu Branch of the People's Bank of China jointly issued the “Implementation Opinions on Establishing a Risk Compensation Fund System for Rural Revitalization and Agricultural Industry Development Loans” (hereinafter referred to as the “Risk Compensation Fund System”). The Risk Compensation Fund System (hereinafter referred to as RCFS) has been gradually implemented throughout Sichuan Province since then. The government directly intervenes in the agricultural credit process while providing credit risk compensation, aiming to reduce financial institutions' risk expectations for agricultural industry loans, help financial institutions share risks, and encourage financial institutions to increase loan issuance to rural revitalization and agricultural industry fields, thereby promoting high-quality agricultural development.

Credit support policies for agricultural industry development in China mainly include targeted reserve requirement ratio cuts and loan interest subsidies. While targeted reserve requirement ratio cuts can effectively improve the loan availability for agricultural enterprises and small and micro enterprises (Guo et al., 2019; Lin et al., 2020; Kong et al., 2021), they mainly support enterprise-type agricultural industry organizations, leaving non-enterprise agricultural and rural economic organizations with insufficient support. Loan interest subsidy policies, despite reducing financing costs are selectively implemented by rural financial institutions due to the significantly higher default probability of agriculture-related loans compared to non-agriculture-related loans (Yin et al., 2014). This has negatively impacted the operating performance of rural financial institutions (Wang and Wen, 2019), leading to the suppression of agriculture-related loans and insufficient support for agriculture after commercialization (Dong et al., 2016; Yang and Yang, 2018). The key issue lies in financial institutions' higher expected risks and insufficient internal incentives to support agriculture. To strengthen policy-based financial support for agriculture, it is crucial to improve the credit risk-sharing mechanism and reduce financial institutions' risk expectations for agriculture-related credit (Wang et al., 2018).

Unlike the two credit support approaches mentioned above, Sichuan Province's RCFS directly intervenes in agricultural credit risk management at the microeconomic level, innovatively altering the risk-sharing mechanism and precisely addressing the critical challenges in financial support for agriculture. The RCFS differs from conventional policy-based agricultural guarantee mechanisms in that it does not compensate for all risk-based loans, with banks continuing to bear a portion of the risk, thus preserving market principles. Simultaneously, it introduces participation from government functional departments and multiple stakeholders in pre-loan review, post-loan supervision, and the recovery of loss loans. Compared to ordinary agricultural loans, the RCFS enhances financial institutions' loan risk management capabilities. Under the participation of government functional departments, commercial banks' loan risk management challenges are effectively shared and reduced within the RCFS framework, which holds positive significance for improving financial institutions' willingness to support agriculture.

As an institutional mechanism innovation policy, Sichuan Province adopted a prudent implementation strategy, stipulating that local governments should implement the system independently according to local conditions, with county-level governments generally formulating specific implementation plans. According to information collected and compiled by this study, 79 districts and counties issued Risk Compensation Fund System implementation documents by the end of 2018, followed by 38, 13, 8, and 7 districts and counties respectively from 2019 to 2022, formally implementing the Risk Compensation Fund System. This policy has gradually expanded throughout the province, and as of late October 2023, it has been implemented for nearly 5 years, representing a typical multi-period policy.

The RCFS differs from general policy-based agricultural guarantee mechanisms. It does not compensate for all risky loans. A portion of the risk is still borne by banks, preserving the market principle. However, it introduces the functions of pre-loan review, post-loan supervision, and recovery of non-performing loans with the participation of government functional departments and multiple entities. Compared with ordinary agricultural loans, the RCFS enhances financial institutions' loan risk management capabilities. The difficulty of commercial banks in handling loan risks is effectively shared and reduced within the framework of the RCFS, which improves financial institutions' willingness to support agriculture.

Despite the potential significance of the RCFS, existing research rarely involves this type of policy, let alone studies on such a unique system. Most studies only evaluate the Average Treatment effect on the Treated (ATT) of policies, while this paper also uses the Synthetic Difference-in-Differences (SDID) method to estimate the Average Treatment Effect on the Untreated (ATU). This paper aims to fill these research gaps by examining how the government micro-intervenes in agricultural credit through the visible hand in the RCFS and its policy effects, using methods such as Staggered Difference-in-Differences (DID), CSDID (Callaway and Sant'Anna, 2021), and Synthetic Difference-in-Differences (SDID). The study also investigates the impact of the RCFS policy design mechanism on policy effects.

The theoretical analysis in this study draws upon the literature on credit support policies, financial system reforms, and the visible hand in agricultural credit. Based on the theoretical analysis, we propose two main hypotheses: H1: The RCFS has a significant positive effect on agricultural development, but there is heterogeneity in the policy effects across counties and time. And H2: County-level bank competition has a significant moderating effect on the relationship between the RCFS and agricultural development. The policy effect is expected to be stronger in counties with higher levels of bank competition.

The empirical analysis in this study employs a comprehensive dataset covering 161 counties in Sichuan Province from 2010 to 2021. The results reveal that the RCFS significantly increases the added value of the primary industry by 14% and promotes agricultural development, supporting H1. The policy effects are heterogeneous across counties and time, with the SDID estimates showing an ATT of 0.068 and an ATU of 0.106. Furthermore, the study finds that county-level bank competition and policy design features, such as the participation of guarantee companies and the leverage ratio of the risk compensation fund, play important roles in shaping the policy's impact, partially supporting H2.

Sichuan Province's RCFS represents an important financial innovation, attempting to address the mismatch between agricultural financial supply and demand through microeconomic matching mechanisms. It moves beyond simple price control or price subsidy policies, such as the widely implemented agricultural credit interest rate controls or agricultural credit interest subsidies. This paper thoroughly examine how the government uses the visible hand to intervene in agricultural credit at the microeconomic level through the RCFS and its policy effects, providing important implications for developing countries to improve their agricultural credit policies. In particular, this research holds strong implications for developing countries under strict fiscal constraints, specifically regarding the optimal allocation of fiscal funds to support agricultural development.

Additionally, this study of the RCFS adds new evidence to understanding the government's role in economic development and contributes microeconomic elements to the optimization of agricultural credit policy mechanism design. More importantly, the RCFS serves as an important case for observing and understanding the ‘visible hand,' and this paper makes contributions to deepening the understanding of the ‘visible hand' and ‘intervention.' Overall, as a unique credit policy, there is almost no existing research on the RCFS, and this paper fills this gap, advancing the research boundaries of agricultural credit policy.

The remainder of this paper is structured as follows: Section 2 reviews the relevant literature, Section 3 describes the implementation measures of the RCFS, Section 4 presents the theoretical analysis framework of the RCFS and research hypotheses, Section 5 empirically analyzes the policy effects and mechanisms of the RCFS, and Section 6 concludes with policy implications.

2 Literature review

Agriculture plays a crucial role in economic development. It contributes to economic growth through food supply (Johnson, 2000), population growth (Johnson, 1997), industrialization (Gollin et al., 2002), and economic structural transformation (Timmer, 2009). Governments have formulated various financial support policies to promote agricultural development, which can be categorized as credit support policies, financial system reforms, and risk management tool support policies.

Credit support policies play a crucial role in fostering agricultural development in financially underdeveloped economies. Historical evidence from the United States reveals that the establishment of Production Credit Associations in the early twentieth century significantly improved the short-term credit availability of farm owners, leading to increased yield and revenue in corn production (Hutchins, 2023). Similarly, empirical evidence from Brazil shows that credit support policies can substantially boost farmers agricultural output (Maia et al., 2019). The mechanisms underlying these positive effects are multifaceted. Credit rationing constrains the realization of economies of scale of land (Zhang et al., 2018), while credit constraints hinder smallholder farmers' adoption of modern agricultural technologies, such as small-scale irrigation, fertilizers, and improved seeds (Balana et al., 2022). Conversely, enhanced credit availability can effectively reduce the incidence of land abandonment (Du et al., 2019) and significantly increase farms' adoption of Integrated Crop-Livestock Systems (Carrer et al., 2020). Moreover, rural credit policies have been shown to substantially improve agricultural productivity (Sakhno et al., 2019; Agbodji and Johnson, 2021; Ogbeide-Osaretin and Aliu, 2022). In the context of China, research has revealed an inverted U-shaped relationship between agricultural credit scale and total factor productivity in the agricultural sector (Wen and Wang, 2022).

Financial system reforms attempt to create a more favorable financial environment for agricultural development from an institutional perspective. Due to information asymmetry (Stiglitz and Weiss, 1981), lack of collateral, and other reasons, agricultural economic organizations often face financial service dilemmas such as “credit rationing” by financial institutions in the financial market (Ma et al., 2011; Li, 2014). To solve this systemic problem, many countries have established financial institutions serving agricultural development. For example, the U.S. cooperative banking system directly provides seasonal and medium-term loans, while Japan has established a sound agricultural cooperative financial system (Lei and Zhang, 2016). Thailand, on the other hand, supports agricultural and rural development with the Bank for Agriculture and Agricultural Cooperatives as the main body (Yang, 2009). The reform of rural credit cooperatives is one of the important reforms in China's financial system (Liu and Xu, 2003), focusing on clarifying property rights and improving the management system (He, 2004). The impact of property rights reform on the efficiency of rural credit cooperatives shows an inverted U-shaped trend (Zhang et al., 2014), but the reform's effect on supporting agriculture has not met expectations (Zhao and Sun, 2010). The establishment of new types of rural financial institutions is also an important financial reform measure (Hong, 2011). Additionally, with the development of financial technology and the Internet, the role of microfinance in agricultural development has also received considerable attention (Wang, 2015; Nakano and Magezi, 2020).

Risk management tool support policies primarily consist of agricultural insurance subsidy policies and insurance product innovation. Agricultural insurance has been shown to significantly increase agricultural output (Wang, 2011) and farmers' income (Zhang and Sun, 2015). Moreover, policy-based insurance serves as a critical prerequisite for the development of new agricultural business entities (Zhang and Zhao, 2013). Consequently, public financial subsidies for agricultural insurance have become a widely adopted policy across countries to develop policy-based agricultural insurance (Xiao et al., 2013). In addition to subsidies, encouraging insurance product innovation, such as weather insurance (Xu et al., 2014) and agricultural product price insurance (Wagener and Zenker, 2021) has also emerged as an essential policy trend to manage agricultural risks, increase production, and safeguard farmers' income.

In the above-mentioned financial policies, the government plays the role of a macro-control decision-maker, subsidizer, and reformer. However, in the RCFS established by Sichuan Province, the government directly participates in the micro-implementation process of agricultural credit, attempting to optimize the allocation of agricultural credit and reduce credit risk expectations through the visible hand. Part of the credit risk is borne by government funds, changing the risk expectations of borrowers and financial institutions. At the same time, the RCFS is neither a fully market-oriented operation nor a completely public financial subsidy policy. In the actual implementation process, the RCFS is not a market intervention behavior, and the main body's willingness of the market is fully retained.

Since the RCFS in Sichuan Province, China, is a very unique credit policy with partial risk retention, i.e., the government not only retains part of the principal risk of agricultural credit but also deeply participates in agricultural credit risk management, existing research rarely involves the RCFS, let alone studies on such a unique RCFS. This paper's important contribution lies in filling this gap, expanding the research boundary of agricultural credit policy, deepening people's understanding of the “visible hand” and “intervention”. Moreover, unlike most studies that actually only evaluate the Average Treatment effect on the Treated (ATT) of policies, this paper also uses the SDID method to estimate the ATU (Average Treatment effect on the Untreated) of this policy, providing a new perspective and new ideas for better comprehensive evaluation and understanding of policy mechanism design.

3 Implementation measures of the risk compensation fund system

3.1 Establishment of the risk compensation fund system

The RCFS implemented by Sichuan Province aims to support the development of agricultural industries and new types of agricultural business entities with growth potential. The policy is implemented independently by district and county governments within the framework of the “Implementation Opinions on Establishing a Risk Compensation Fund System for Rural Revitalization and Agricultural Industry Development Loans,” with each formulating local policies based on their specific conditions. Special teams are established to manage the RCFS, incorporating different government departments and financial institutions.

The organizational structure involves collaboration among the Finance Bureau, Agriculture and Rural Affairs Bureau, township government, and local People's Bank of China. Each department has specific responsibilities, such as fund management, technical support, loan application review, and credit supervision.

3.2 Operating mechanism

The RCFS operating mechanism involves division of labor and collaboration among multiple departments, as shown in Supplementary Figure 1.

The financing process begins with the entity submitting a loan application to the competent authority. The application undergoes review by the village committee and/or township government before being transferred to the cooperating bank. The bank's opinion determines loan approval, ensuring market-driven supply and demand.

In some cases, an agricultural guarantee company is involved, bearing partial risk. The competent authority conducts an expert evaluation, and the guarantee company issues a formal guarantee letter if approved. The bank then issues the loan, with some districts requiring additional review.

During the loan use stage, the financial institution conducts post-loan management, while the competent authority and township government supervise the entity's loan use. Upon completion, the entity must repay the loan on schedule. In case of default, the financial institution and government departments collaborate to recover funds and minimize losses.

4 Theoretical analysis framework of the risk compensation fund system promoting agricultural development mechanism

4.1 Theoretical analysis

The RCFS changes the agricultural credit risk-sharing mechanism in the overall credit process, enhances financial institution' willingness to increase agricultural credit supply, and promotes agricultural development through multiple channels. Supplementary Figure 2 summarizes the theoretical mechanism of the RCFS in promoting agricultural development, highlighting the key components and their interactions.

The RCFS promotes agricultural development by optimizing the customer selection mechanism, risk control mechanism, and credit allocation mechanism of agricultural credit. These mechanisms are grounded in the literature on credit support policies (Maia et al., 2019; Hutchins, 2023), financial system reforms (Liu and Xu, 2003; He, 2004), and the visible hand in agricultural credit (Wang et al., 2018).

First, the RCFS sets entry thresholds, forming a selection mechanism for credit customers. Those thresholds are based on factors such as operational soundness, scale, and credit records, forming a customer selection mechanism that directs credit resources toward agricultural entities with growth potential (Zhang et al., 2018; Du et al., 2019). The RCFS policy targets new agricultural business entities with development potential, clearly stating that the financing entities must have stable operations and good credit records. Cooperatives, family farms, and large-scale farmers must be registered and have been in actual operation for more than 1 year. Cooperatives must reach a certain scale, and the number of cooperative members should not be less than 50 in principle. Other agricultural and rural economic organizations with appropriate scale operations are also within the scope of support. The deep participation of multiple government departments enables cooperating banks and agricultural guarantee companies to more effectively select and tap potential credit customers, reduce financial institutions' customer acquisition costs, improve financial institutions' credit efficiency and financing entities' credit availability, and provide sufficient credit resources for agricultural development.

Second, the RCFS has a relatively precise credit allocation mechanism, further optimizing the matching mechanism between credit supply and credit demand in the rural financial market, channeling funds to key agricultural projects and entities (Wang and He, 2019; Mo and Shen, 2020). The RCFS' credit allocation mechanism is designed to optimize the matching between credit supply and demand in the rural financial market. It takes into account factors such as the type of agricultural entity, the scale of operation, the intended use of funds, and the potential for growth and development. By directing credit resources toward key agricultural projects and entities that have the potential to drive rural economic growth, the RCFS aims to maximize the impact of agricultural credit on rural development.

Third, the “non-intervention” participation of multiple government departments and entities in the RCFS improves the credit risk control mechanism, reducing financial institutions' risk perceptions and enhancing their willingness to lend to the agricultural sector (Yang and Yang, 2018; Wang and Wen, 2019). This collaborative approach combines administrative and market resources to reshape the rural financial service system (Lu, 2003). The RCFS's credit risk control mechanism involves the participation of multiple government departments and entities, such as the Agriculture and Rural Affairs Bureau, the Finance Bureau, and the People's Bank of China. These departments work together to assess the creditworthiness of financing entities, monitor the use of funds, and provide technical support and guidance. By sharing information and resources, they can more effectively identify and mitigate credit risks, reducing financial institutions' risk perceptions and enhancing their willingness to lend to the agricultural sector.

Fourth, the RCFS can improve financial institutions' risk tolerance. By improving financial institutions' risk tolerance, the RCFS can stimulate the entrepreneurial spirit of financing entities, promoting innovation and growth in the agricultural sector (Nakano and Magezi, 2020). The RCFS provides a risk compensation mechanism that shares the potential losses from agricultural loans. This risk-sharing arrangement can encourage financial institutions to be more willing to lend to agricultural entities, even those with higher risk profiles. As a result, financing entities may have greater access to credit and be more willing to invest in new technologies, expand their operations, and take on more ambitious projects, thereby stimulating entrepreneurship and innovation in the agricultural sector.

Fifth, by supporting the development of new agricultural business entities, the RCFS can generate spillover effects on agricultural development, driving the growth of agricultural industrial chains and rural economies (Zhang and Zhao, 2013; Lei and Zhang, 2016). These entities often have close ties to smaller farmers and can serve as models for modern agricultural practices and technologies. By providing credit support to these entities, the RCFS can help them expand their operations, improve their efficiency, and increase their competitiveness. This, in turn, can drive the growth of agricultural industrial chains and rural economies, as these entities create new jobs, stimulate demand for agricultural products and services, and contribute to rural income growth.

The RCFS can promote agricultural development through the above theoretical mechanisms. However, the policy effects may exhibit heterogeneity due to differences in natural conditions, economic development levels, government financial conditions, and agricultural development conditions among regions (Arkhangelsky et al., 2021). For example, some district and county documents directly stipulate that when the non-performing loan ratio or bad debt ratio of the RCFS reaches a certain standard, the RCFS will be suspended, the source of risk will be thoroughly investigated, loan losses will be cleared, and the policy can continue to be implemented only after that. Such differences in risk control conditions set by the competent government departments can result in heterogeneity in the policy effects of the RCFS.

Furthermore, county-level bank competition is a concentrated embodiment of the operating conditions of the rural financial market and an important factor driving financial institutions to participate in the rural financial market. It has an important impact on inclusive financial services, credit fund flow, and farmers' income (Wang and He, 2019; Mo and Shen, 2020; Wang and Zhao, 2023). The policy effect of the RCFS on agricultural development is subject to the macro influence of county-level bank competition. County-level bank competition simultaneously affects credit efficiency and credit availability, playing a regulating role in the policy effect of the RCFS.

4.2 Research hypotheses

Based on the above theoretical mechanisms of the RCFS influencing agricultural development, we propose the following research hypotheses:

H1: The RCFS has a significant positive effect on agricultural development, but there is heterogeneity in the policy effects across counties and time.

H2: County-level bank competition has a significant moderating effect on the RCFS and agricultural development. The policy effect is expected to be stronger in counties with higher levels of bank competition.

5 Empirical analysis

5.1 Data description

The data used in this paper consist of two parts: policy variables and main economic variables at the county level. The policy variable, which is the core explanatory variable, indicates whether a district or county has implemented the RCFS system by the end of 2021. The binary variable takes a value of 1 if the district or county has implemented the RCFS and 0 otherwise. This policy variable data is collected from government websites, telephone inquiries to the Agriculture and Rural Affairs Bureau of the district or county, and inquiries to the mailboxes of county-level government leaders, such as county mayors, district mayors, and county party secretaries. The establishment and implementation date of the RCFS are based on the documents issued by the government or official replies.

The second part of the data comprises the main economic variables of counties in Sichuan Province, obtained from the county statistical yearbooks spanning from 2010 to 2021. The selected economic variables primarily focus on agriculture and overall fiscal and economic conditions. The added value of the primary industry is used as the main variable to measure agricultural development. The statistical characteristics of the data are shown in Table 1.

Table 1

Table 1. Statistical characteristics of data.

Table 2 gives the inter-group mean difference test results for major economic variables between counties with and without RCFS implementation before the policy treatment. The results suggest that there are no significant differences in agricultural development between the two groups, while there are disparities in economic and fiscal strength. These findings imply a low likelihood of confounding variables and reverse causality when estimating the RCFS's policy effect on agricultural development.

Table 2

Table 2. Mean difference test of major economic variables between counties with and without policy implementation.

5.2 Benchmark model

To estimate the effect of the RCFS on agricultural development, this paper uses the Staggered Difference-in-Differences (DID) as the benchmark model, given the staggered implementation of the policy across districts and counties. The RCFS is promoted in different districts and counties at different time periods. The specific year of policy implementation varies among different districts and counties, making it a multi-period policy. The RCFS was officially introduced in Sichuan Province at the end of 2018, with different districts and counties implementing the policy in 2019, 2020, and 2021. As the actual implementation and deployment of the policy began in 2019, even for districts and counties that issued policy documents at the end of 2018, 2018 should be regarded as the period before policy treatment.

The benchmark model is shown in Equation 1.

${\text{Y}}_{\text{it}}={\lambda }_0+{\text{D}}_{\text{it}}\theta +{\alpha }_{\text{i}}+{\lambda }_{\text{t}}+{\sigma }_{\text{it}}(1)$

In Equation 1, Y_it is the dependent variable of agricultural development. In the empirical analysis, Y_it is the logarithmic value of the added value of the primary industry. α_i and λ_t control for individual fixed effects and year fixed effects, respectively. D_it is the policy variable of the RCFS, which varies with the district, county, and year. θ is the policy effect of the RCFS. σ_it is the random disturbance term. i represents the i-th district or county, and t represents the year.

The empirical analysis excludes districts and counties for which official information on the establishment of the RCFS is unavailable. The final sample consists of 161 districts and counties in China's Sichuan Province, forming a panel data-set spanning from 2010 to 2021.

5.3 Benchmark model regression results

Table 3 presents the regression results of the benchmark model. Columns 1 to column 4, control for county and year fixed effects, and robust standard errors are clustered at the county level to account for potential within-county correlation. Column 1 does not include control variables, while columns 2, 3, and 4 include control variables. Columns 2 and 3 separately include the ratio of local fiscal general budget expenditure to local fiscal general budget revenue and the ratio of local fiscal general budget expenditure to regional GDP, respectively. Column 4 includes both control variables in the benchmark regression model. These control variables are added to address potential confounding factors that may simultaneously affect agricultural development and the policy implementation of the RCFS.

Table 3

Table 3. Regression results of the baseline model.

Across all specifications, the policy coefficient of the RCFS is positive and statistically significant at the 1% level. The benchmark model regression results show that the implementation of the RCFS is associated with a 14.2% increase in the added value of the primary industry. The magnitude and significance of this result remain stable when adding the control variables in columns 2 to 4, with the policy coefficient ranging from 0.140 to 0.143. These results indicate that the RCFS promotes agricultural development, supporting research hypothesis 1.

5.4 Parallel trend test

To verify the parallel trend necessary for the difference-in-differences (DID) analysis and RCFS before implementation, this paper refers to the event study method proposed by Callaway and Sant'Anna (2021). Supplementary Figure 3 presents the results of this test. The horizontal axis represents the policy treatment period, with period 0 being the initial policy treatment period. Supplementary Figure 3 shows that before the actual implementation of the policy in 2019, the estimated policy effects are not statistically distinguishable from zero. Following implementation, the policy exhibits a statistically significant positive effect at the 5% level. These findings support the parallel trend assumption, suggesting the observed policy impacts are attributable to the RCFS.

5.5 Robustness test

5.5.1 Replacing the dependent variable

The impact of the RCFS on agricultural development may be multifaceted. To assess the robustness of the RCFS' impact on agricultural development, we conduct regressions with alternative dependent variables: the logarithmic values of grain output, oil crop output, and meat output. All specifications control for county and year fixed effects, employing clustered standard errors at the county level. Table 4 presents the regression results, where columns 2, 4, and 6 include all control variables in the benchmark model. The RCFS' coefficent remains significantly positive for oil crop output and meat output, supporting the RCFS' role boosting production of these key agricultural products.

Table 4

Table 4. Regression results with replaced dependent variables.

5.5.2 CSDID estimation results

The RCFS is a typical multi-period policy, which can result in biases in the staggered DID analysis due to incorrect group comparisons. To address potential biases for multi-period policies, we further evaluate the RCFS's impact using the CSDID method proposed in Callaway and Sant'Anna (2021). This approach estimates the overall policy treatment effect for groups receiving the policy at different times.

The dependent variable of CSDID is still the logarithmic value of the added value of the primary industry. The estimation results show that the overall ATT of the RCFS is 0.0751, with a standard error of 0.0289. Policy effects for 2019, 2020, and 2021 treatment groups are 0.0317, 0.0736, and 0.1198, respectively. The CSDID-estimated ATT is significant at the 5% level, verifying the robustness of the RCFS's policy effect. However, the overall treatment effect of the policy is lower than the estimation result of the benchmark model, suggesting potential differences in how these methods capture policy effects.

5.5.3 Bacon decomposition

Goodman-Bacon (2021) demonstrates that staggered DID estimates for multi-period policies is a weighted average of effects across various group comparisons between the control group, early treatment group, and late treatment group. Biases may arise if the early treatment group is used as the reference group for the late treatment group. To obtain an unbiased policy treatment effect, we apply Bacon decomposition using primary industry added value as the dependent variable. The decomposition results are shown in Table 5.

Table 5

Table 5. Bacon decomposition results.

The Bacon decomposition yields a policy effect of 0.141 for the RCFS. Interestingly, the comparison between the late and early treatment groups receive the smallest weight. The Bacon decomposition result aligns with our benchmark model, indicating that the regression results of the benchmark model are robust. Collectively, the results of replacing the dependent variable and the CSDID regression results, support H1 regarding the RCFS' positive impact on agricultural development.

5.6 Policy dynamic effects of different groups

To investigate the short and long-term impacts of the RCFS, which was implemented in batches across Sichuan Province, we employ the method of Callaway and Sant'Anna (2021). We employ the logarithmic value of the added value of the primary industry as the dependent variable, and estimates the dynamic effects of the treatment groups from 2019 to 2021, as shown in Supplementary Figure 4.

Our analysis reveals a time lag of approximately 1 year before the RCFS exerts significant effects. Initial implementation shows non-significant effects across groups. However, a positive trend emerges in the second year, with the most significant effects observed in the third year for groups treated in 2018 and in 2019. While the 2020 treatment group exhibits benefits in 2021, these are not statistically significant.

These findings strongly suggest that the RCFS is a long-term mechanism rather than a short-term policy mechanism that can produce immediate results. Policymakers should prioritize the long-term development of agriculture over short-term benefits, allowing adequate time for trial and error, and policy adjustment.

5.7 Individual treatment effects and heterogeneity analysis

Considering that the RCFS is implemented independently by local district and county governments, there are large differences among districts and counties in terms of natural geographical conditions, economic development levels, fiscal conditions, and agricultural development conditions. The policy effects of the RCFS may also vary greatly among different districts and counties. Then, how are the policy effects in the districts and counties that have implemented the RCFS? Although the analysis results of the benchmark model and robustness tests show that the average treatment effect on the treated (ATT) of the RCFS in Sichuan Province is significant and can promote agricultural development, it is difficult to determine based on this that the RCFS is effective in all treated districts and counties. Have all the implementing districts and counties achieved the desired policy effects? Do districts and counties with higher economic development levels have better policy implementation effects? Do districts and counties with better terrain conditions have better policy effects? Further asking, if the policy were implemented in those non-implementing control group districts and counties, what would be the policy effect (ATU)? If the ATT and ATU of the RCFS can be obtained simultaneously, then the ATE can be further calculated to re-evaluate the overall implementation effect of the RCFS in Sichuan Province. This can answer whether the policy should be promoted throughout the province and is an important basis for the expansion of the RCFS.

Recognizing potential variation across districts and counties in Sichuan Province (e.g., geography, economic development), we investigated the RCFS's impact at a granular level. While the benchmark model and robustness tests suggest a positive average treatment effect on the treated (ATT), the extent to which this applies to all treated districts and counties remains unclear. Additionally, we seek to estimate the policy's potential effect on non-implementing control districts and counties (ATU) to inform broader policy expansion.

To address these questions, this paper uses the Synthetic Difference-in-Differences (SDID) method to estimate the individual treatment effect (ITE) of each district and county (Arkhangelsky et al., 2021). The SDID method combines the DID method with the Synthetic Control Method (SCM). Unlike the Synthetic Control Method, which only assigns different weights to different individuals (Abadie et al., 2010), the SDID also assigns weights to time variables, possessing double robustness. First, we use the SDID method to estimate the individual treatment effects (ITE) for treated districts and counties. We then treat the original control group as if they had implemented the RCFS in 2019, estimating their ITEs. The individual treatment effects of the control group districts and counties are then estimated one by one, i.e., assuming that these districts and counties that have not actually implemented the policy have implemented the RCFS, with the districts and counties that have already implemented the policy serving as the control group to synthesize the districts and counties that have virtually implemented the policy.

According to the estimation results of the Synthetic DID, the ATT of the RCFS is 0.068, the ATU is 0.106, and the ATE is 0.075. This estimation result is close to the CSDID estimation result and lower than the Staggered DID estimation result. The ATU of 0.106 suggests substantial potential for the RCFS in currently untreated districts and counties. This highlights the policy's positive marginal benefit in Sichuan Province and supports its broader implementation across Sichuan province.

We further match the ITE of each district and county with the average slope data of terrain conditions and the per capita GDP data in 2018, as shown in Supplementary Figure 5. Despite substantial heterogeneity in outcomes across different terrain conditions and economic development levels, effective policy implementation was observed in both favorable and less favorable contexts. This indicates that the RCFS should be implemented even in areas with geographical or economic challenges, emphasizing the importance of the policy's own design and mechanisms. The policy does not necessarily obtain greater policy benefits in districts and counties with superior terrain conditions and economic development levels. From this perspective, the RCFS still has a large policy space to continue its implementation in Sichuan Province.

Moreover, compared with external conditions, the policy mechanism itself may be more important. Although the basic operational mechanism of the RCFS is broadly similar across all implementing counties in Sichuan Province, there are significant variations in its practical implementation because each county and district carries out the policy autonomously. These differences are manifested in several key aspects: the structure of the RCFS fund pool, the designated supervisory authority, and the guarantee support system.

First, regarding the structure of the RCFS fund pool. There are three primary models. In one category of counties, the RCFS fund pool is established and capitalized solely by the county-level finance bureau. In a second model, several county-level finance bureaus jointly contribute funds to establish a single, larger RCFS fund pool supervised by the agriculture and rural affairs bureau of the prefecture-level city. This pooled fund is then used collectively by the participating counties, with the prefecture-level bureau responsible for daily management and approvals. Ya'an City, for example, has adopted this approach, which allows counties with less favorable geographical or economic conditions to access a fund pool with a greater scale of capital and stronger fiscal support. A third category combines the first two, where RCFS fund pools exist and operate concurrently at both the prefecture-level city and county levels. For instance, the county-level RCFS might handle credit demands below a certain threshold (e.g., 2 million RMB), while larger demands are managed by the prefecture-level RCFS. Yibin City employs this third type of hybrid fund pool structure.

Second, there are variations in the supervisory authority for the RCFS. In most counties, the RCFS is supervised directly by the county's agriculture and rural affairs bureau. However, some county bureaus delegate the management and daily operations of the RCFS to a local agricultural financing guarantee company. Although these guarantee companies in China are state-capitalized, they operate according to market-based principles. Consequently, an RCFS managed by such a company tends to exhibit greater market-oriented operational efficiency.

Third, the guarantee support systems upon which the RCFS relies also differ. In China, the guarantee support system fors agricultural development is structured across three tiers: provincial, prefecture-level city, and county-level agricultural financing guarantee companies. However, not all prefecture-level cities or counties have established their own local companies. This leads to a situation where some counties, when implementing the RCFS, can only rely on the participation of the provincial-level company. In contrast, other counties benefit from the support of not only the provincial and municipal-level companies but also a locally established county-level company. For instance, Yibin City has access to the Sichuan provincial company, the Yibin municipal company, and, within one of its districts, the Xuzhou District Agricultural Financing Guarantee Co., Ltd.

Therefore, these three sources of variation—in fund pool structure, supervisory authority, and the depth of the guarantee system—will lead to differential policy effects of the RCFS across counties. They represent the primary mechanisms explaining why the RCFS can produce strong positive effects even in counties with less favorable geographical or economic conditions.

5.8 Mechanism Analysis

5.8.1 Regulating effect of county-level bank competition

We employ the method of Chong et al. (2013) to calculate the Herfindahl-Hirschman Index (HHI) for different districts and counties as a proxy for competition. The smaller the value of HHI, the greater the degree of competition among county-level banks. We analyze the regulating effect of county-level bank competition on the policy effect by creating an interaction term between HHI and the policy variable D.

Table 6 gives the regression results of the regulating effect of county-level bank competition. From columns 1–4, the dependent variables are the logarithmic values of the added value of the primary industry, grain output, oil crop output, and meat output, respectively. County fixed effects and year fixed effects are controlled, and standard errors are clustered at the county level. The regression results in Table 6 show that after controlling for bank competition, the policy's main effect becomes largely insignificant. This indicate that the policy's effectiveness may depend on financial resources and competition levels. Except for column 3, interaction terms between the policy variable and HHI are significantly positive, indicating competition regulate policy's impact. However, these positive coefficients imply that less competition leads to greater policy effectiveness, partially contradicting H2.

Table 6

Table 6. County-level banking competition moderating effects.

This result challenges Hypothesis H2, which posited that the policy effect would be stronger in counties with higher bank competition. However, from a theoretical perspective, this finding may reflect the dual-edged nature of bank competition in imperfect financial markets, particularly in rural China where agricultural lending is characterized by high information asymmetries and default risks. In counties with lower competition (higher HHI), banking markets tend to exhibit oligopolistic features, enabling institutions to engage in “relationship lending” that fosters long-term ties with borrowers. According to Petersen and Rajan (1995), concentrated markets incentivize banks to invest in soft information gathering, reducing adverse selection and moral hazard, which in turn enhances the efficiency of policy interventions like the RCFS's risk-sharing mechanism. This is especially relevant in agricultural contexts, where banks in less competitive environments are more likely to align with government directives, channeling funds effectively to high-risk sectors rather than diverting them to urban or non-agricultural opportunities (Cetorelli and Gambera, 2001).

Conversely, in highly competitive counties, banks face greater profit pressures and risk aversion, potentially weakening the policy's marginal impact. Intense competition can lead to “cherry-picking” of low-risk borrowers, diluting targeted agricultural credit as institutions prioritize short-term returns over policy compliance (Stiglitz and Weiss, 1981). Beck et al. (2006) further argue that in unevenly developed financial systems—like those in many Sichuan counties—excessive competition exacerbates tensions between market incentives and policy goals, resulting in subdued transmission of credit support. Our positive interaction term thus aligns with evidence from developing economies, where moderate market concentration can amplify directed lending policies (Carlson et al., 2022), while hyper-competition introduces frictions that undermine them (Ang, 2008).

Overall, this contradictory outcome enriches the heterogeneity analysis by highlighting bank market structure as a key moderator of RCFS effectiveness. It suggests that H2 holds conditionally in more mature markets but not universally in rural settings, informing policymakers to adapt interventions—such as bolstering incentives in competitive areas—to optimize outcomes.

5.8.2 Analysis of policy mechanism design

The RCFS includes several key mechanisms: participation of agricultural guarantee companies, the amplification multiple of the total loan amount, and the risk proportion borne by the risk compensation fund. We investigate whether variations in these mechanisms across districts and counties correlate with differences in policy effectiveness.

Some districts and counties have included agricultural guarantee companies in the RCFS, allowing guarantee companies to bear a certain proportion of risks for loans issued by the risk compensation fund, while some districts and counties have not included guarantee companies in the RCFS. For the amplification multiple of the total loan amount of the RCFS, most districts and counties have chosen to set it at 10 times, and there are few cases where it is amplified to more than 10 times. For the proportion of risk borne by the risk compensation fund, i.e., when the loans issued within the framework of the RCFS are unrecoverable and form bad debt losses, the risk compensation fund needs to compensate according to a certain proportion. Most districts and counties have chosen a bearing proportion of 30%, and only a few districts and counties have expanded the compensation proportion to above 30%.

Supplementary Figure 6 presents the distribution of policy effects in terms of guarantee participation, amplification multiple, and bearing proportion. Supplementary Figure 6 reveals that the participation of agricultural guarantee companies does not necessarily enhance the policy implementation effect, and the larger the loan amplification multiple of the RCFS does not result in a greater policy implementation effect. Additionally, the policy implementation effects of the risk compensation fund's bearing proportion at 30% and 40% are superior to the policy implementation effects at 50% and above. Therefore, the policy implementation effect of the RCFS does not show a linear relationship with guarantee participation, the amplification multiple of the risk compensation fund, and the bearing proportion. This means that merely increasing fiscal resources does not guarantee the achievement of desired outcomes and may instead lead to inefficient use of fiscal funds and increased loan risks. There is still a large room for optimization in the policy mechanism design of the RCFS.

6 Conclusions and policy implications

This paper first uses the Staggered Difference-in-Differences (DID) method to estimate the overall policy effect of the RCFS and obtains the benchmark regression results. Our results show that the implementation of the RCFS can increase the added value of the primary industry by around 14%. Robustness tests further confirm the effectiveness of the policy, indicating that the policy can promote agricultural development, and research hypothesis H1 is verified. The analysis of the dynamic effects of different groups shows that the districts and counties that implemented the policy earlier have obtained policy benefits. Meanwhile, it also indicates that the policy effect of the RCFS is relatively lagging, and it is difficult to produce immediate policy effects in the short term. The implementation effect of the RCFS in districts and counties with different economic development levels and terrain conditions shows obvious dispersion characteristics. On the whole, the policy benefit of continuing to promote the implementation of the policy in the entire Sichuan Province is positive. County-level bank competition has a significant regulating effect on the policy effect of the RCFS. However, bank competition does not necessarily enhance the policy implementation effect, and research hypothesis H2 is partially verified. Similarly, unilaterally increasing the overall leverage ratio of the risk compensation fund does not necessarily lead to more optimal policy implementation effects.

The findings reveal the importance of well-designed government intervention in agricultural development, common in many developing countries. The RCFS addresses risk expectations and organizational constraints in agricultural lending, enhancing credit allocation efficiency. The micro-intervention of the visible hand solves the allocation problem of agricultural credit, effectively promoting agricultural development, although the general equilibrium effect may not be ideal.

In light of these robust findings, we propose the following policy implications: First, The RCFS is a valuable model for stimulating agricultural growth, warranting further promotion and implementation. Second, the RCFS can succeed in less favorable geographic or economic conditions when tailored to local realities. Third, financial system reform policies promoting county-level bank competition may not align perfectly with rural revitalization and agricultural development goals. Fourth, optimizing fund use and refining policy implementation mechanisms are more crucial than simply increasing fiscal input. Finally, micro-level considerations and targeted government intervention within a market framework are essential for effective agricultural credit policies in developing countries.

While this paper provides empirical evidence for understanding the effects of government intervention in agricultural credit, it is important to acknowledge its limitations. The primary limitation is the time scope of our data. Due to the staggered rollout of the policy and our panel data ending in 2021, the effective post-treatment observation period for any given county is limited, ranging from 1 to 4 years. Consequently, our findings constitute a robust short- to medium-term evaluation of the RCFS during its critical initial implementation phase. We must explicitly caution that these results should not be extrapolated to infer the long-term viability or sustainability of the policy, as its effects may evolve or diminish over a longer horizon. Other limitations include the focus on Sichuan Province, which may affect generalizability, and the emphasis on macro-level outcomes.

Therefore, the primary avenue for future research is a direct follow-up study to assess the long-term impacts of the RCFS. Such an analysis is essential for determining whether the significant short-to medium-term effects documented in this paper persist over time. Future work could also expand the geographical scope and employ qualitative methods to explore the micro-level responses of agricultural organizations to the policy.

Data availability statement

The data supporting the findings of this study are available from the corresponding author upon reasonable request. Requests to access these datasets should be directed to d2FuZ2ZlaWZlaUBzdXNlLmVkdS5jbg==.

Author contributions

XS: Conceptualization, Formal analysis, Funding acquisition, Methodology, Writing – original draft. JL: Resources, Supervision, Writing – review & editing. FW: Project administration, Writing – review & editing. LL: Writing – review & editing.

Funding

The author(s) declare that financial support was received for the research and/or publication of this article. This work was supported by the National Social Science Fund Project, “Research on the Docking Mechanism and Reform Path for Improving the Quality and Efficiency of Rural Financial Services in Rural Revitalization” (Grant No. 19CJY032) and the Sichuan Science and Technology Program, “Study on the Operating Mechanism, Model Comparison, and Optimization Path of the Loan Risk Compensation Fund System for Agricultural Industry Development in Rural Revitalization from the Perspective of High-Quality Development” (Grant No. 2023JDR0130). Both grants were awarded to author S. Su. The funding bodies provided financial support for the research project but had no role in the study design; in the collection, analysis, or interpretation of data; in the writing of the manuscript; or in the decision to submit the article for publication.

Conflict of interest

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Generative AI statement

The author(s) declare that Gen AI was used in the creation of this manuscript. We disclose that generative AI (Chat GPT 4o by Open AI) was used to help edit this manuscript for language polishing and clarity.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher's note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Supplementary material

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fsufs.2025.1592138/full#supplementary-material

References

Abadie, A., Diamond, A., and Hainmueller, J. (2010). Synthetic control methods for comparative case studies: estimating the effect of California's tobacco control program. J. Am. Stat. Assoc. 105, 493–505. doi: 10.1198/jasa.2009.ap08746

Crossref Full Text | Google Scholar

Acemoglu, D., and Robinson, J. A. (2013). Why Nations Fail: The Origins of Power, Prosperity, and Poverty. Reprint edition. New York: Crown Currency.

Google Scholar

Agbodji, A. E., and Johnson, A. A. (2021). Agricultural credit and its impact on the productivity of certain cereals in togo. Emerg. Mark. Fin. Trade 57, 3320–3336. doi: 10.1080/1540496X.2019.1602038

Crossref Full Text | Google Scholar

Ang, J. B. (2008). What are the mechanisms linking financial development and economic growth in Malaysia?. Econ. Model. 25, 38–53. doi: 10.1016/j.econmod.2007.04.006

Crossref Full Text | Google Scholar

Arkhangelsky, D., Athey, S., Imbens, G. W., and Wager, S. (2021). Synthetic difference-in-differences. Am. Econ. Rev. 111, 4088–4118. doi: 10.1257/aer.20190159

Crossref Full Text | Google Scholar

Balana, B. B., Mekonnen, D., Haile, B., Hagos, F., Yimam, S., and Ringler, C. (2022). Demand and supply constraints of credit in smallholder farming: evidence from Ethiopia and Tanzania. World Dev. 159:106033. doi: 10.1016/j.worlddev.2022.106033

Crossref Full Text | Google Scholar

Beck, T., Demirgüç-Kunt, A., and Levine, R. (2006). Bank concentration, competition, and crises: First results. J. Bank. Fin. 30, 1581–1603. doi: 10.1016/j.jbankfin.2005.05.010

Crossref Full Text | Google Scholar

Callaway, B., and Sant'Anna, P.H.C. (2021). Difference-in-Differences with multiple time periods. J. Econ. 225, 200–230. doi: 10.1016/j.jeconom.2020.12.001

Crossref Full Text | Google Scholar

Carlson, M., Correia, S., and Luck, S. (2022). The effects of banking competition on growth and financial stability: evidence from the national banking era. J. Polit. Econ. 130, 462–520. doi: 10.1086/717453

Crossref Full Text | Google Scholar

Carrer, M. J., Maia, A. G., de Mello Brandão Vinholis, M., and de Souza Filho, H. M. (2020). Assessing the effectiveness of rural credit policy on the adoption of integrated crop-livestock systems in Brazil. Land Use Policy 92:104468. doi: 10.1016/j.landusepol.2020.104468

Crossref Full Text | Google Scholar

Cetorelli, N., and Gambera, M. (2001). Banking market structure, financial dependence and growth: international evidence from industry data. J. Fin. 56, 617–648. doi: 10.1111/0022-1082.00339

Crossref Full Text | Google Scholar

Chong, T. T-. L., Lu, L., and Ongena, S. (2013). Does banking competition alleviate or worsen credit constraints faced by small- and medium-sized enterprises? Evidence from China. J. Bank. Fin. 37, 3412–3424. doi: 10.1016/j.jbankfin.2013.05.006

Crossref Full Text | Google Scholar

Dong, X., Zhou, L., and Liu, J. (2016). Selective formulation and implementation of financial support for agriculture policy—also on policy at the top, countermeasures at the bottom. Issues Agricult. Econ. 37, 18–30. doi: 10.13246/j.cnki.iae.2016.10.003

Crossref Full Text | Google Scholar

Du, J., Zeng, M., Xie, Z., and Wang, S. (2019). Power of agricultural credit in farmland abandonment: evidence from rural China. Land 8:184. doi: 10.3390/land8120184

Crossref Full Text | Google Scholar

Goldman, E., and Gorton, G. (2000). The Visible Hand, the Invisible Hand and Efficiency'. National Bureau of Economic Research (Working Paper Series). New York, NY: National Bureau of Economic Research.

Google Scholar

Gollin, D., Parente, S., and Rogerson, R. (2002). The role of agriculture in development. Am. Econ. Rev. 92, 160–164. doi: 10.1257/000282802320189177

Crossref Full Text | Google Scholar

Goodman-Bacon, A. (2021). Difference-in-differences with variation in treatment timing. J. Econ. 225, 254–277. doi: 10.1016/j.jeconom.2021.03.014

Crossref Full Text | Google Scholar

Guo, Y., Xu, F., and Shu, Z. (2019). The Inclusive Effect of Targeted Reserve Requirement Ratio Reduction Policy Under the Background of Banking Competition—Based on the Analysis of A-share and NEEQ Agricultural, Rural and Micro Small Enterprises Data. Fin. Res. 1–18.

Google Scholar

He, G. (2004). Problems and Solutions in the Reform of Rural Credit Cooperatives. Bankers, 38–39.

Google Scholar

Hong, Z. (2011). Is the reform of new rural financial institutions feasible?—An analysis from the perspective of supervision efficiency. Econ. Res. J. 44–58.

Google Scholar

Hutchins, J. (2023). The US farm credit system and agricultural development: Evidence from an early expansion, 1920–1940. Am. J. Agricult. Econ. 105, 3–26. doi: 10.1111/ajae.12290

Crossref Full Text | Google Scholar

Johnson, D. G. (1997). Agriculture and the wealth of nations. Am. Econ. Rev. 87, 1–12.

Google Scholar

Johnson, D. G. (2000). Population, Food, and Knowledge. Am. Econ. Rev. 90, 1–14. doi: 10.1257/aer.90.1.1

Crossref Full Text | Google Scholar

Keech, W. R., and Munger, M. C. (2015). The anatomy of government failure. Public Choice. 164, 1–42. doi: 10.1007/s11127-015-0262-y

Crossref Full Text | Google Scholar

Kong, D., Li, H., and Yang, W. (2021). Targeted reserve requirement ratio reduction, loan accessibility, and commercial credit of micro and small enterprises—empirical evidence based on regression discontinuity. Fin. Res. 77–94.

Google Scholar

Krueger, A. O. (1990). Government failures in development. J. Econ. Perspect. 4, 9–23. doi: 10.1257/jep.4.3.9

Crossref Full Text | Google Scholar

Lei, D., and Zhang, X. (2016). Experience and Enlightenment from the United States and Japan on rural financial support for agricultural modernization. Rural Fin. Res. 50–54. doi: 10.16127/j.cnki.issn1003-1812.2016.05.010

Crossref Full Text | Google Scholar

Li, S. (2014). Financial service dilemmas and optimization paths for family farms—based on 151 family farms in Guangxi. Southern Fin. 62–66.

Google Scholar

Lin, C., Chen, X., and Wang, H. (2020). Study on the precision of targeted reserve requirement ratio reduction benefiting agriculture based on enterprise micro perspective. China Rural Observ. 773, 83–101.

Google Scholar

Liu, M., and Xu, Z. (2003). Rural Credit Cooperative Reform and Government Functions. J. Econ. (Q.) 555–572.

Google Scholar

Lu, L. (2003). Reshaping the three-level rural financial service system with administrative and market resources. Fin. Res. 106–114.

Google Scholar

Ma, D., Du, J., Shi, C., and Li, G. (2011). Empirical analysis of credit financing and growth of specialized farmers' cooperatives in underdeveloped areas—based on the survey of demonstrative farmers' professional cooperatives in Gansu Province. Chinese Rural Econ. 34–41. doi: 10.20077/j.cnki.11-1262/f.2011.07.004

Crossref Full Text | Google Scholar

Maia, A. G., Eusébio, G. dos S., and da Silveira, R. L. F. (2019). Can credit help small family farming? Evidence from Brazil. Agricult. Fin. Rev. 80, 212–230. doi: 10.1108/AFR-10-2018-0087

Crossref Full Text | Google Scholar

Mo, Y., and Shen, K. (2020). County bank branch layout and rural credit fund flow. J. South China Agricult. Univ. 11–23.

Google Scholar

Nakano, Y., and Magezi, E. F. (2020). The impact of microcredit on agricultural technology adoption and productivity: evidence from randomized control trial in Tanzania. World Dev. 133:104997. doi: 10.1016/j.worlddev.2020.104997

Crossref Full Text | Google Scholar

Ogbeide-Osaretin, E. N., and Aliu, T. I. (2022). Empirical investigation on domestic credit to private sector and rural development in Nigeria. J. Dev. Econ. 4, 35–46. doi: 10.47672/jde.1150

Crossref Full Text | Google Scholar

Petersen, M. A., and Rajan, R. G. (1995). The effect of credit market competition on lending relationships. Q. J. Econ. 110, 407–443. doi: 10.2307/2118445

Crossref Full Text | Google Scholar

Sakhno, A., Kharlamova, G., Kushnir, N., and Hrechanyuk, O. (2019). Impact of credit and investment resources on the productivity of agricultural sector. Eur. J. Sustain. Dev. 8, 335–335. doi: 10.14207/ejsd.2019.v8n2p335

Crossref Full Text | Google Scholar

Stiglitz, J. E., and Weiss, A. (1981). Credit rationing in markets with imperfect information. Am. Econ. Rev. 71, 393–410.

Google Scholar

Timmer, P. C. (2009). A World Without Agriculture: The Structural Transformation in Historical Perspective. Washington, DC: Aei Press.

Google Scholar

Wagener, A., and Zenker, J. (2021). Decoupled but not neutral: the effects of counter-cyclical cash transfers on investment and incomes in rural Thailand†. Am. J. Agricult. Econ. 103, 1637–1660. doi: 10.1111/ajae.12172

Crossref Full Text | Google Scholar

Wang, J., Xiao, Q., and Li, J. (2018). Financing of new agricultural management entities: dilemmas, causes, and countermeasures—based on the survey of 131 agricultural comprehensive development industrialization development loan interest subsidy projects. Issues Agricult. Econ. 71–77. doi: 10.13246/j.cnki.iae.2018.02.008

Crossref Full Text | Google Scholar

Wang, W., and Wen, T. (2019). Did agricultural loans drag down the operating performance of rural financial institutions?. Agricult. Technol. Econ. 73–84. doi: 10.13246/j.cnki.jae.2019.02.006

Crossref Full Text | Google Scholar

Wang, X. (2011). The impact of agricultural loans and agricultural insurance on agricultural output—evidence from prefecture-level units in China from 2004 to 2009. Chin. Rural Econ. 44–51. doi: 10.20077/j.cnki.11-1262/f.2011.10.005

Crossref Full Text | Google Scholar

Wang, X. (2015). Research on internet finance helping solve the financing difficulty of “long tail” micro and small enterprises. Fin. Res. 128–139.

Google Scholar

Wang, X., and He, G. (2019). County banking competition and the deepening of inclusive financial services—stratified analysis of poor and non-poor counties. Chin. Rural Econ. 55–72. doi: 10.20077/j.cnki.11-1262/f.2019.04.005

Crossref Full Text | Google Scholar

Wang, X., and Zhao, Y. (2023). County banking competition and common prosperity of farmers—dual perspectives of absolute income and relative income. Econ. Res. J. 98–115.

Google Scholar

Wen, H., and Wang, H. (2022). Spatial-temporal relationship between agricultural credit scale and agricultural total factor productivity. Modern Econ. Discuss. 114–125. doi: 10.13891/j.cnki.mer.2022.04.011

Crossref Full Text | Google Scholar

Xiao, W., Luo, B., and Wang, K. (2013). Public finance subsidizing agricultural insurance: international experience and practice in China. Chin. Rural Econ. 13–23. doi: 10.20077/j.cnki.11-1262/f.2013.07.002

Crossref Full Text | Google Scholar

Xu, S., Jin, L., and Dong, M. (2014). Practice and Experience of Foreign Agricultural Weather Risk Management Financial Innovation Products. World Agricult. 25–29.

Google Scholar

Yang, G. (2009). International Comparative Study on Government Support for Rural Finance. 25, 42–46.

Google Scholar

Yang, Y., and Yang, S. (2018). Study on the effectiveness of rural commercial banks supporting agriculture from the perspective of target deviation—based on data from 107 rural commercial banks in Hunan Province. China Soft Sci. 42–53.

Google Scholar

Yin, Z., Xie, H., and Wei, Z. (2014). Agricultural Loans, Monetary Policy and Default Risk. Chin. Rural Econ. 14–26. doi: 10.20077/j.cnki.11-1262/f.2014.03.002

Crossref Full Text | Google Scholar

Zhang, H., Luo, J., and Niu, R. (2014). Property rights reform and efficiency change of rural credit cooperatives and its convergence: 2008-2014—evidence from 107 counties (Districts) in Shaanxi Province. Manag. World 92–106. doi: 10.19744/j.cnki.11-1235/f.2017.05.009

Crossref Full Text | Google Scholar

Zhang, L., Feng, Y., Wang, W., and Zhao, R. (2018). Agricultural scale economy and land productivity under credit rationing. Chin. Rural Econ. 19–33.

Google Scholar

Zhang, X., and Sun, R. (2015). Regional differences in the impact of agricultural insurance on farmers' income—based on panel data cluster analysis. Insur. Stud. 62–71. doi: 10.13497/j.cnki.is.2015.06.006

Crossref Full Text | Google Scholar

Zhang, Z., and Zhao, H. (2013). Overcoming the dilemmas of new agricultural management entities and system and mechanism innovation. Reform 78–87.

Google Scholar

Zhao, L., and Sun, W. (2010). Policy effect assessment of rural credit cooperative reform on improving financial support for agriculture—based on national rural household survey panel data DID analysis. Fin. Res. 194–206.

Google Scholar

Appendix

Table A1

Table A1. Regression results with D_new.