Introduction

Rural areas in South Korea are facing a serious extinction crisis due to factors such as population decline, aging, and economic stagnation (Kim, 2024). Of the 228 cities and counties, 118 are considered at risk of extinction, meaning over 50% of these areas are threatened, with Jeollabuk-do having 13 out of 14 cities and counties at risk and Gangwon-do having 16 out of 18 (Lee et al., 2024). This highlights the relative vulnerability of rural areas compared to urban regions, emphasizing the need to enhance the competitiveness of agricultural industries to overcome these challenges (Kim and Lee, 2014).

The Ministry of Agriculture, Food and Rural Affairs has implemented various agricultural and food support programs in response to the vulnerabilities of rural areas, covering sectors such as agricultural production infrastructure, rural communities, food, and distribution. In particular, a significant focus has been placed on policies aimed at restructuring the income structure of farming households, and these efforts have led to a steady increase in farmers’ incomes (Jung et al., 2013; Kim and Hwang, 2009). According to the “Farm Household Economy Survey” conducted by Statistics Korea in 2023, the average farm household income nationwide was 50.82 million KRW, with non-farm income accounting for 19.99 million KRW or 39.3% of the total (Kosis, 2023). However, the primary reason for the increase in farm household income is the rise in transfer income, such as subsidies and direct payments (Kim et al., 2016). While this increase in transfer income may temporarily boost household income, it cannot serve as a long-term solution for stabilizing farm household income. Therefore, for the improvement in farm household income to lead to an enhancement in the standard of living of farmers, strategies are needed that go beyond traditional methods such as income preservation and focus on maximizing the value of economic activities within rural areas (Hwang and Lee, 2016).

In this context, the Rural Convergence Industry (RCI) Policy can be seen as an active and cyclical approach that adds value to agriculture by connecting farming with industries such as rural tourism, processing and selling local agricultural products, and more (Joo, 2020). Specifically, the RCI refers to activities that create added value by integrating the primary (agriculture), secondary (processing and development of agricultural products), and tertiary (food service and hospitality) industries (Kim and Heo, 2011; Yang et al., 2014). Through the RCI, it is possible to revitalize local agricultural industries, encourage rural return migration, and improve the conditions for sustainable development in rural areas (Lee, 2017). In the context of a declining rural population and shrinking regional industries, exploring ways to strengthen local competitiveness is necessary, and the RCI Policy could serve as a suitable alternative for revitalizing regional economies, including rural areas.

Therefore, this study aimed to analyze the policy effects of the RCI Policy on regional economies. To achieve this, data on the number of RCI-certified businesses, representing the implementation of the RCI Policy, was used as a key variable, while the agricultural and fisheries gross regional domestic product (GRDP), a comprehensive indicator of the agricultural economy in the region, was used as the outcome variable. The RCI certification system was established by the “Act on the Promotion and Support of Rural Convergence Industry” (hereafter referred to as the Rural Convergence Industry Act) on June 3, 2014, and implemented on June 4, 2015, operating under Article 8 (Certification of Rural Convergence Industry Operators).

However, the number of RCI-certified businesses is not randomly determined, which raises concerns about endogeneity. Therefore, instrumental variables were used to address the endogeneity issue and enhance the reliability of the results. In this study, considering the characteristics of the data, the panel two-stage least squares (2SLS) method was applied, with the number of rural experience and recreation villages and the number of local food direct sales stores, serving as instrumental variables.

Review of Previous Studies

Review of Previous Studies

The RCI strategy combines manufacturing, processing, sales, and experience functions with agriculture to increase the income of agricultural management entities (Kim, 2016). Through the RCI, added value can be created, which has become a major source of non-farm income and a strategy to diversify the income sources of rural residents (Park, 2013). Furthermore, the RCI creates new jobs in rural areas, increasing employment in various sectors such as processing and tourism (Kim, 2013). However, despite the positive effects of the RCI on regions, economic analysis regarding its impact on the activation of the regional agricultural economy is still lacking.

In a study analyzing the relationship between the RCI and regional development, Yoo and Ryu (2015) examined the current status and issues of the RCI through in-depth interviews and case studies, offering insights into how the value of agriculture and rural areas can be enhanced. Specifically, they emphasized the need for RCI businesses to establish solid infrastructure and highlighted that managers must grow beyond farming, processing, and sales to become food experts.

Chung et al. (2016) presented specific policy examples and the current status, along with the trend that the RCI Policy aims to create overall jobs and added value in rural areas. In particular, they highlighted the observed increase in the profitability of agricultural corporations and farming households, as well as the job creation effects through the RCI. They also emphasized the importance of private-sector-led partnerships for revitalizing the local economy and forming value chains.

Kim et al. (2017) conducted a study focused on the RCI districts promoted by the RCI Act. The study found that the designation of these districts, driven by district support projects, led to sequential conflicts during the planning, evaluation, and operation stages, hindering the smooth implementation of the projects. They proposed solutions and improvements for each stage to address these issues.

Additionally, regarding the income-increasing effects of the RCI, Park et al. (2014) analyzed the impacts on agricultural income and non-farm income by using data from the 2010 Agricultural, Forestry, and Fishery Census and the 2011 Agricultural Corporation Survey, based on different types of RCI combinations. The study found that the combination of primary and secondary industries (1 × 2) or primary and tertiary industries (1 × 3) was more likely to generate high incomes than the combination of primary, secondary, and tertiary industries (1 × 2 × 3). They suggested the need for a partial implementation of RCI strategies as a policy recommendation.

Hwang and Lee (2016) analyzed the selection tendencies of RCI types based on the 2010 Basic Survey on Agricultural and RCI, considering the characteristics of different groups. They also categorized the RCI into comprehensive and single types to determine which type was more effective in increasing income. The results, consistent with the findings of Park et al. (2014), suggested that single convergence is more suitable than comprehensive convergence for Korea’s agricultural environment. They recommended that policies encouraging comprehensive convergence may not be appropriate for Korea.

As such, while previous studies analyzing the relationship between the RCI and regional revitalization have suggested directions for the future of the RCI, they have primarily been qualitative and lacked concrete numerical support. Research on the relationship between the RCI and income involved quantitative analysis, but it focused more on increasing farm household income rather than the overall development of the region. Furthermore, a limitation is that only short-term data from the early stages of the RCI Policy was used, making it difficult to definitively assess the overall effectiveness of the policy.

This study aims to advance existing research by economically analyzing the effects of the RCI on the revitalization of the regional agricultural economy, using the number of certified RCI operators and the agricultural, forestry, and fisheries GRDP as key variables. While previous studies did not thoroughly address the relationship between the number of certified RCI operators and the regional economy, this research offers a new perspective in this regard. Additionally, by utilizing panel data from nine provinces spanning from 2015 to 2022, this study considers regional characteristics and seeks to assess the effects of the policy not only in its early stages but also in more recent years. A distinctive feature of this study is its use of instrumental variables to address the endogeneity issue associated with the number of certified RCI operators, thus enhancing the validity of the analysis. This research design is expected to contribute to identifying strategies for the sustainable development of the RCI and the regional agricultural economy.

Research Methodology

Data Selection

This study aimed to analyze the impact of the RCI Policy on the revitalization of the regional agricultural economy. The study covered the period from 2015, when the RCI Policy was initiated, to 2022, the most recent year for which the GRDP data are available, spanning eight years. The analysis included panel data from nine provinces: Gyeonggi, Gangwon, Chungcheongbuk, Chungcheongnam, Jeollabuk, Jeollanam, Gyeongsangbuk, Gyeongsangnam, and Jeju. The RCI was managed by local support centers in each province, and these nine provinces serve as key hubs for the RCI in South Korea.

In this study, the dependent variable and variable of interest used were the agricultural and forestry GRDP and the number of RCI-certified operators. The GRDP represents the sum of the value of final goods and services produced within a certain region over a specific period (Yang et al., 2004). It is a key indicator of regional economic performance, with a higher GRDP signifying a higher level of economic development in the region (Kim, 2010; Kim and Choi, 2022). Therefore, this study used the agricultural and forestry GRDP as the dependent variable to comprehensively assess the regional industry and economic development in the agricultural and forestry sectors. Additionally, the number of RCI-certified operators, which serves as a key variable representing the implementation of the RCI, played a role in evaluating the policy’s effectiveness in this study.

The control variables in this study primarily included agricultural production factors, farm income, and the structure of farming households. These variables are representative factors that influence the varying effects of the policy across regions. By controlling for them, a more accurate evaluation of the impact of the RCI on the regional economy can be achieved (Kim et al., 2012).

First, the farm population and cultivated area are key indicators of production factors. The farm population and cultivated area are the foundation of agriculture, directly linked to productivity, and they directly influence the efficiency and profitability of agriculture (Han, 2015). In agricultural productivity research, farm population by age and gender and cultivated area by size are essential variables (Kim and Lee, 2000; Kim et al., 2019; Kim et al., 2020).

Farm income is a key variable that directly benefits from the RCI. Moreover, the increase in farm income can lead to higher local consumption and an improvement in the quality of life, making it closely related to the agricultural and forestry GRDP (Kim and Chai, 2009; Yoon, 2008). In this study, reflecting the fact that a significant portion of the total farm income comes from agricultural production and agricultural-related activities, agricultural income and non-agricultural income were treated as separate variables for the analysis.

Additionally, differences in the farm structure can lead to varying levels of participation in and effectiveness of the RCI. Generally, smaller-scale farms or those with older managers experience reduced productivity, and the aging of participants in the RCI can hinder the enhancement of participants’ capabilities, creating a structure that makes it difficult to foster development through innovation (Sim and Chung, 2016; Woo and Kim, 2016). Furthermore, farms with a higher level of digitalization are better positioned to utilize the latest information and agricultural technology through online platforms and mobile applications, which enhances their accessibility to markets and consumers (Hwang and Kim, 2023). This becomes a factor that can maximize participation in the RCI and the effectiveness of the policies. For this reason, this study included older adult farmers, small-scale farmers, and digitally advanced farmers as control variables to reflect the farm structure. To account for multicollinearity, these variables were converted into ratios relative to the total number of farms. In this context, older adult farmers refer to those whose farm managers are 65 years of age or older, small-scale farmers refer to those with less than 1 hectare of cultivated land, and digitally advanced farmers refer to those who use computers, laptops, or smartphones.

This study also used instrumental variables to account for the endogeneity of the variable of interest. An instrumental variable is considered appropriate when it is not correlated with the error term, is strongly related to the variable of interest, and has no direct or only an indirect effect on the dependent variable (Min, 2008). In other words, the appropriate instrumental variables in this study must be closely related to the number of certified RCI operators but only have an indirect effect on the agricultural and forestry GRDP. From this perspective, the study used the number of rural experience and leisure villages and the number of local food direct sales outlets as instrumental variables.

The RCI is a concept that integrates agriculture, manufacturing, and services to create added value. Rural experience and leisure villages combine agricultural experiences with tourism, providing opportunities to process and experience local agricultural products. Local food direct sales outlets directly sell fresh agricultural products produced in the region to consumers, offering small-scale farms a promotional platform while raising local residents’ awareness of regional agricultural products (Jang et al., 2013; Jin and Chae, 2013; Park, 2013). These aspects are closely related to the manufacturing and service sectors of the RCI.

Specifically, Table 1 presents the selection criteria for RCI-certified businesses. The certification process includes an assessment of essential infrastructure for operating an RCI business, such as the presence and management of agricultural product processing facilities, experience centers, and direct sales outlets. Additionally, as part of the qualitative evaluation, applicants are assessed on how they plan to integrate and utilize local direct sales outlets and experience centers in connection with their agricultural activities. Therefore, obtaining RCI certification implies that the business meets the standards related to rural experience and leisure villages, as well as local food direct sales outlets, indicating a strong correlation between the number of RCI-certified businesses and the number of rural experience and leisure villages and local food direct sales outlets.

These two facilities also satisfied the conditions between the instrumental variables and the dependent variable. The number of rural experience and leisure villages and local food direct sales outlets does not directly affect the agricultural, forestry, and fisheries GRDP; any potential impact would occur only when these facilities are utilized for manufacturing and service industries, leading to product output. This implies that the number of rural experience and leisure villages and local food direct sales outlets influences the agricultural, forestry, and fisheries GRDP indirectly through the RCI.

A higher number of these facilities might be expected to have a positive (+) impact on the agricultural, forestry, and fisheries GRDP. However, local food direct sales outlets and rural experience and leisure villages are currently not operating effectively. In the case of local food direct sales outlets, their quantitative growth has not been accompanied by a reflection of consumer needs (Yu and Um, 2022). Additionally, farmers’ participation in store operations and agri-food sales remain low, and these outlets lack a system to promptly restock sold-out items (Cho and Park, 2016; Jeong et al., 2018). Rural experience and leisure villages often meet only the formal criteria while losing their sense of community due to declining or changing village populations, with many being operated by only a small number of individuals (Yi et al., 2021). The necessary foundation to enhance customer satisfaction and loyalty in these villages has not yet been established (Hwang and Lee, 2020). These prior studies support the argument that the mere number of rural experience and leisure villages and local food direct sales outlets does not directly impact the agricultural, forestry, and fisheries GRDP.

Analytical Model

This study considered the issue of endogeneity between the variable of interest and the dependent variables, namely the number of RCI-certified businesses and the agricultural, forestry, and fisheries GRDP. The causal relationship between these two variables is not clearly defined, and unobserved factors beyond the number of RCI-certified businesses may also influence the agricultural, forestry, and fisheries GRDP. Such unobserved variables can introduce bias into the results. In this context, the instrumental variable approach is employed to mitigate endogeneity issues. To estimate the causal relationship between the independent and dependent variables, this study adopted the 2SLS method, which utilizes instrumental variables as an alternative estimation technique. The dataset used in this study comprised panel data spanning eight years across nine provinces. Accordingly, a panel 2SLS analysis was conducted.

Panel analysis is a quantitative method of analyzing panel data, comprising both time-series and cross-sectional data (Kim, 2014). One of its advantages is that it can control for estimation errors arising from time-series processes as well as those from regional-level data (Jeong et al., 2013). Panel models are classified into the fixed effects model (FEM) and the random effects model (REM). When selecting the appropriate model for this study among the FEM, REM, and pooled ordinary least squares (POLS), standard tests such as the F-test, Breusch–Pagan Lagrange multiplier (LM) test, and Hausman test were conducted. The results demonstrated that the fixed effects model was the most suitable, leading to the final adoption of the FE-2SLS model.

The linear regression model assuming the error term ui as a fixed effect in this study is expressed in Equation (1). In this equation, A.GRDP represents the agricultural, forestry, and fisheries GRDP; N6I denotes the number of RCI-certified businesses; FS refers to the farm structure (digitalized farms, farms with older adult farmers, and small-scale farms); Prod indicates production factors (farm population and cultivated land area); and FI represents farm income (agricultural and non-agricultural income). Since cov(N6I,ϵit)≠0 in Equation (1), the 2SLS estimation using instrumental variables is required to obtain a consistent estimator.

First, the within estimator was obtained by setting the endogenous explanatory variable N6I as the dependent variable and using the instrumental variables—the number of rural experience villages (Z1it) and the number of local food direct stores (Z2it)—along with the exogenous variables FS, Prod, FI as explanatory variables (Min and Choi, 2012). This constituted the first stage of the 2SLS estimation, as shown in Equation (2).

From Equation (2), N6I^ can be derived. By substituting N6I^ for N6I in Equation (1) and re-estimating, the FE-2SLS estimator can be obtained. This constituted the second stage of the 2SLS estimation, which is summarized in Equation (3).

Results and Discussion

Descriptive Statistics

The descriptive statistics for the variables used in this study are presented in Table 2 below. The average number of RCI-certified businesses was found to be 169, and the average agricultural and forestry GRDP was calculated at 360 trillion 6.028 billion KRW. Additionally, the proportion of information-oriented farms was 42.8%, while the proportion of farms with older adult farmers was 57.4%, showing a relatively balanced distribution compared to farms without older adult farmers. The proportion of small-scale farms was 69.5%, indicating a relatively high share. The farm population and cultivated land area were 236,790 individuals and 170,190 ha, respectively, and non-agricultural income was higher than agricultural income at 17.07 million KRW.

Panel Model Estimation Results

The panel model was divided into FEM and REM. To determine the most appropriate model for this study, the characteristics of the entities need to be considered. Based on the F-test, the estimate was 43.30, which was rejected at the 1% significance level. The LM test showed an estimate close to zero, indicating that we could not reject the null hypothesis. This suggests that the FEM is more suitable than the POLS model, while the POLS model is more suitable than the REM. Therefore, the fixed effects model was considered the most appropriate model for this study.

Table 3 below presents the results of the general FEM. According to the analysis, for each unit increase in the number of RCI-certified operators, the agricultural and fisheries GRDP increases by 1.71 billion KRW. Additionally, the agricultural and fisheries GRDP increases by 7 million KRW for each unit increase in the farm population, 15 million KRW for each unit increase in the arable land area, and 29 million KRW for each unit increase in off-farm income. These results showed similar significance levels to those obtained from the subsequent panel 2SLS estimation, but since the endogeneity of the variable of interest was not fully controlled, they were likely underestimated.

Panel 2SLS Estimation Results

Based on the test results of the panel model, a fixed-effects model was applied for the panel 2SLS estimation. The analysis results are presented in Table 4 below. In the first stage of the panel 2SLS estimation, the number of RCI-certified operators was set as the dependent variable, and the control variables along with the instrumental variables, such as the number of rural experience and recreation villages and the number of local food direct sales stores, were used as explanatory variables. The results showed that the ratio of small-scale farms and off-farm income was significantly analyzed at the 10% level, while the number of rural experience villages and local food direct sales stores were significantly analyzed at the 5% and 1% levels, respectively. This indicates that small-scale farms or farms with low off-farm income are more likely to participate in the RCI Policy, and the number of rural experience villages and local food direct sales stores are significantly related to the number of RCI-certified operators as instrumental variables.

The dependent variable in the second stage of the panel 2SLS estimation was the agricultural and forestry GRDP. The estimation results showed that as the number of RCI-certified operators increases, the agricultural and forestry GRDP increases by 3.1 billion KRW. This indicates that the RCI Policy has a significant positive effect on the local agricultural and forestry economy. Specifically, it suggests that as the number of RCI-certified operators increases, the linkage between industries strengthens. This result reflects the expected effects of the policy, such as job creation and income growth. According to the “2021 RCI Basic Status Survey Report” published by the Ministry of Agriculture, Food and Rural Affairs and the Korea Rural Community Corporation, there are 104,067 RCI businesses, with 320,000 employees, and the revenue of certified businesses exceeds 23 trillion KRW. This indicates that certified businesses outperform non-certified ones in terms of sales and employment, aligning with the findings of this study.

However, the analysis also indicated that as the proportion of small-scale farms increases, the agricultural and forestry GRDP decreases. This result reflects the fact that small-scale farms, compared to large-scale farms, tend to have relatively low productivity and lack sufficient technology and capital. Previous studies such as that conducted by Woo and Kim (2016) also pointed out that small-scale farms typically exhibit lower productivity and profitability.

The finding that as the rural population and cultivated land area increase, the agricultural and forestry GRDP also rises suggests that productivity in agriculture increases as more labor and land are invested into farming. This is consistent with studies on agricultural productivity, such as those by Han (2015) and Kim et al. (2020), which focused on the relationship between agricultural labor and land inputs and productivity. Additionally, in terms of income, only off-farm income was analyzed to be significant. This can be attributed to the fact that off-farm income accounts for approximately 40% of the total farm household income, and various policies have been implemented to support off-farm income, which is expected to have influenced these results. This is consistent with the findings of Park et al. (2014) and Hwang and Lee (2016), who argued that the RCI can enhance farmers’ income, suggesting that this policy may contribute to increasing the economic stability of farm households.

Meanwhile, in this study, the Davidson–Mackinnon exogeneity test was conducted to verify the endogeneity of the number of RCI-certified operators. Generally, in panel 2SLS models, the Hausman test is commonly conducted, which is applicable when the difference in the covariance matrix of the estimated coefficients tends to be positive asymptotically (Choi, 2014; Min and Choi, 2012). In this study, however, the difference in the covariance matrix was -3.26, indicating a negative sign. The Davidson–Mackinnon exogeneity test serves as an alternative method in such cases. The test result showed a p-value of 0.0678, rejecting the null hypothesis at the 10% level. Therefore, it can be concluded that endogeneity exists in the number of RCI-certified operators.

Summary and Conclusion

This study was conducted to analyze the effect of the RCI Policy on the regional agricultural economy. The study period spanned from 2015 to 2022, with panel data obtained from nine provinces: Gyeonggi, Gangwon, Chungcheongbuk, Chungcheongnam, Jeollabuk, Jeollanam, Gyeongsangbuk, Gyeongsangnam, and Jeju. The number of RCI-certified operators was used as the key variable, while the agricultural and forestry GRDP, a comprehensive indicator of the regional agricultural economy, was used as the outcome variable. Considering the potential endogeneity issue between these two variables, the 2SLS method was chosen as the research model, utilizing instrumental variables. Given the panel data structure, the panel 2SLS method was applied. The instrumental variables used were the number of rural experience villages and the number of local food direct sales stores.

First, a panel test was conducted to determine which model, the FEM or the REM, was more appropriate, and the FEM was proven to be suitable. Therefore, the FE-2SLS was selected as the final model for this study. In the estimation, the new estimate of the number of RCI-certified operators, (N6I)̂, was found to have a positive effect of 3.1 billion KRW on the agricultural and forestry GRDP. This result indicates that the RCI is making a positive contribution to agriculture-related economic activities. Specifically, it suggests that the integration of industries centered on agriculture has created added value in agriculture, leading to diversified income sources for farmers and increased job creation based on this. These results indicate that the initial objectives of the RCI policy have been achieved. However, contrary to the first-stage estimation, which indicated that a larger proportion of small-scale farms participate in the RCI Policy, the second-stage estimation found that as the proportion of small-scale farms increases, the agricultural and forestry GRDP decreases. This result indicates that while the participation of small-scale farms enhances the policy’s effects, their low productivity limits their economic contribution.

Based on these results, the following policy implications can be drawn:

First, additional discussions should take place to promote inter-industry convergence. To achieve this, the establishment of relevant platforms that enable farmers to collaborate with agricultural corporations and businesses, benchmarking of relevant policies, and the development of cooperation models are recommended. This would provide a foundation of opportunities for farmers to participate in the RCI. Consequently, a more positive impact on the development of the regional agricultural economy may be observed.

Second, practical support is needed to strengthen the foundation of small-scale farms. Currently, although small-scale farms participate in the RCI certification system, their economic contribution remains low. This is because small-scale farms are relatively lacking in resources and technological capabilities and cannot fully expect cost-saving effects through economies of scale. As of 2022, small-scale farms account for 73.4% of all farms in South Korea. When small-scale farms establish a stable production and income base, the stability of regional agriculture can also be secured. Therefore, strengthening the foundation of small-scale farms is essential, and one possible approach is to consider organizing farms. This would help increase the scale and expertise of farms, ensuring a stable distribution and supply system. For farms that face difficulties in organizing due to geographical conditions, policies should be developed to enable local agricultural cooperatives that can serve as hubs.

The RCI is a key strategy for enhancing the productivity and competitiveness of regional agriculture while promoting agricultural sustainability. This study analyzed the impact of the RCI Policy on regional agriculture, confirming its positive contribution to the regional agricultural economy. In this process, instrumental variables were used to address the endogeneity issue between the RCI-related variables and agricultural and forestry GRDP, ensuring the study’s significance as it estimated the practical effects with reduced bias. However, the process of becoming an RCI-certified operator involves various factors, and there may be omitted variables that were not accounted for in this study. Additionally, the RCI involves different types of participation, such as production-oriented and processing-oriented models, which were not considered in this study. This represents a limitation, and future research should address these aspects to provide a deeper understanding of the effectiveness of the RCI Policy.