Agro-Tourism Integration and County-Level Sustainability: Mechanisms and Regional Heterogeneity in China

Abstract

The agro-tourism integration model combines agricultural production, rural life, and tourism experiences, bringing new opportunities for the sustainable development of counties. The objective of this study is to explore the impact of the agro-tourism integration on the sustainable development of counties, reveal the underlying mechanisms and regional heterogeneity effects, and provide a scientific basis for the formulation of rural revitalization policies. Based on panel data from 1749 counties in China from 2008 to 2021, an empirical test using a propensity score matching-difference-in-differences (PSM-DID) model found that the National Policy on Leisure Agriculture and Rural Tourism Demonstration Counties significantly enhances the level of sustainable development in counties through three mechanisms: stimulating consumer demand, attracting capital inflows, and optimizing industrial structure. Moreover, the effect is more pronounced in the western regions of China. In terms of theory, this study shifts the focus to the county level, constructing a comprehensive measurement index system for county-level sustainable development. It analyzes the mechanisms through which the integration of agriculture and tourism operates and verifies the policy effects. In terms of practice, policy implementations are proposed to adopt a multi-pronged approach to increase agricultural-tourism consumption; to promote government-enterprise cooperation and introduce long-term funding; and to develop the service industry in a location-specific manner to continuously optimize the industrial structure. This study indicates that agro-tourism integration is an effective path for the sustainable development of counties. The policy design needs to take into account the differences in regional resource endowments. This has important implications for promoting county-level development in a location-specific manner under the rural revitalization strategy.

1. Introduction

As China’s aging population intensifies and economic and population growth slow down, the “siphon effect” of large cities has led to increasing pressure on capital and human resource outflows in China’s county regions. The development of agro-tourism integration offers opportunities to mitigate or even reverse these trends. Agro-tourism integration is a novel industrial model that combines tourism experiences, agricultural output, natural views, and countryside living. It can be traced back to mid-19th-century Europe, where it primarily catered to the rural tourism needs of the nobility. The growing desire for nature and rural life among people in developed countries has fueled the swift progress of agro-tourism integration [1], with more and more countries promoting the integration of agriculture and tourism through policy support and industry regulation.

In recent years, under the strategic background of rural revitalization and the integration of rural industries, China has also advocated the deep fusion of agriculture and tourism to drive regional development. China has developed a series of characteristic agritourism projects, emphasizing the deep fusion of the three industries based on distinctive agriculture, such as the “Hani Rice Terraces” in Yuanyang County in Yunnan and the “World Jasmine Expo Garden” in Jianwei County in Sichuan. With the increasing demand for natural, healthy, and cultural experiences among urban tourists, agro-tourism integration has brought new opportunities for the sustainable progress of China’s county areas.

The theory of industrial integration holds that the integration between different industries brings new economic growth and more development opportunities [2]. The integration of agro-tourism has demonstrated significant economic and social value in practice and research at the provincial level. In terms of economic value, agro-tourism integration directly promotes the sales of local agricultural products, providing farmers with additional sources of income [3]. It also stimulates regional economic growth and strongly propels the process of rural development [4]. Regarding social value, it creates a large number of job opportunities, effectively reducing unemployment rates [3]. The rise of smart agro-tourism has spurred innovation in related technologies [5]. The traditional agricultural culture and rural customs of rural areas are promoted nationwide with the flow of tourists [6], contributing to the preservation and continuation of traditional culture.

The above studies reflect that agro-tourism integration can bring multifaceted support to regional development, providing an important theoretical basis and practical evidence for this research. However, there are still some shortcomings. First, although the economic and social values brought by agro-tourism integration have been discussed separately, whether this model can bring comprehensive sustainable development in the three aspects of economy, society, and environment still deserves further exploration.

Second, most existing studies focus on the influence of agro-tourism integration on the growth of provinces, neglecting the exploration at the county level. In terms of industrial distribution, China’s agricultural tourism resources are mostly concentrated at the county level [7]. From 2010 to 2024, the proportion of county-level GDP in the national GDP decreased from 50.9% to 38.5% [8]. From 2010 to 2020, the permanent resident population in counties decreased by approximately 37.82 million people, and from 2020 to 2024, the outflow of county population continued to intensify, with the scale of net outflow expanding [8]. The total nitrogen and phosphorus emissions from agricultural sources have long exceeded 50% of the national total emissions, and environmental pollution remains a serious issue [8]. Therefore, sustainable development in China’s counties currently faces significant challenges. Clarifying whether and how agro-tourism integration can strengthen the sustainable growth of counties is crucial for China’s rural revitalization and coordinated urban–rural development.

Third, because of the notable disparities in growth objectives and resource endowments at different levels of regions, the sustainable development measurement index systems constructed in prior research at the provincial and city levels [9,10] are difficult to use to reflect the degree of sustainable development in county-level regions. Therefore, there is a need to improve the sustainable development evaluation indicator system tailored to the features of county-level development.

Identifying the research gaps outlined above, this study formulates three research questions: 1. What is the impact of agro-tourism integration on the sustainable development of counties? 2. What are the mechanisms through which agro-tourism integration affects the sustainable development of counties? 3. Is there a regional difference in the policy effect of the agro-tourism integration? To answer these questions, this study establishes a comprehensive measurement index system for the sustainable development of counties and explores the impact and mechanisms of the representative policy of agro-tourism integration—the national leisure agriculture and rural tourism demonstration county policy—on county-level sustainable development. And a regional heterogeneity test was conducted. The results show that the implementation of the agro-tourism integration policy can enhance the level of sustainable development in counties. This impact is achieved by stimulating consumer demand, attracting capital inflows, and optimizing the industrial structure, with more pronounced effects in western regions.

The academic contributions can be summarized as follows: With regard to perspective, this study shifts the research focus to the county level, constructs a measurement index system for county-level sustainable development, and improves the construction of regional sustainable development index systems. With regard to theory, this study thoroughly analyzes the mechanisms through which agro-tourism integration enhances the level of sustainable development of counties: consumer demand, capital inflows, and industrial structure. This expands the theoretical framework for research on the mechanisms of agro-tourism integration. This is the manifestation of the industrial integration theory in the integration of the primary and tertiary industries. With regard to methodology, this study regards the national policy on leisure agriculture and rural tourism demonstration counties as a “Quasi-Natural Experiment” and uses the propensity score matching-difference-in-differences (PSM-DID) method to identify the effect of policy implementation on the sustainable development of counties. This contribution provides theoretical references for local governments to promote sustainable development through agro-tourism integration.

The following chapters of this study are arranged as follows: Section 2 shows the literature background and hypotheses, Section 3 shows the methodology and data, Section 4 shows the results of the empirical study, Section 5 shows the discussion of the results, and Section 6 shows the policy implications and conclusions.

2. Literature Background and Hypotheses

2.1. Literature Review

2.1.1. County-Level Sustainable Development

Globally, sustainable development is typically characterized as “development that meets the needs of the present without compromising the ability of future generations to meet their own needs” [11]. Specifically, from the dimensions of energy and economy, sustainable development refers to improving energy efficiency, increasing renewable energy, protecting the environment, and promoting comprehensive social and economic progress [12]. From the social dimension, sustainable development emphasizes social equity and inclusiveness [13]. In the process of urban renewal, the concept of sustainable development emphasizes the balance between economic growth, social development, and environmental carrying capacity and ecosystem health in urban development [14,15,16]. From the standpoint of regional growth, the sustainable development concept emphasizes the balanced economic development of regions and the coordination between economic growth and ecological balance [16,17].

Existing studies often use the entropy weight method (EWM) to formulate an indicator system to calculate the degree of sustainable development in regions. For example, 108 statistical indicators were adopted to calculate the sustainability degree of the European Union region [18]. In China, within the provincial context, relevant scholars have established a sustainable development index system covering 3 secondary indicators and 49 tertiary indicators to assess the sustainable development level of China’s 31 provinces, autonomous regions, and municipalities [9]. For a single province in China, scholars have constructed a comprehensive system including metrics such as the GDP growth rate, per capita GDP, employment rate, education level, medical insurance level, energy consumption intensity, and pollutant emissions from three angles: economy, society, and environment, to evaluate the sustainable development level of Shanxi Province [19]. At the municipal level, scholars have evaluated the sustainable development potential of 17 cities from three aspects: ecological supply and demand, ecological well-being performance, and ecological carrying capacity structure [20].

In summary, the essence of regional sustainable development lies in the integrated and equilibrated development of the economy, society, and environment. Existing research has developed measurement indicators for sustainable development at the provincial and municipal levels, but the construction of sustainable development indicators on the county level is still not adequate. Therefore, this study will start from three perspectives: economy, society, and environment, and comprehensively consider the development foundation of Chinese counties and the availability of data to build a comprehensive measurement system for the sustainable development of counties.

2.1.2. Agro-Tourism Integration and Sustainable Development of Counties

Over the past few years, scholars have conducted extensive discussions on the dimensions of the factors that influence the sustainable development of regions. In terms of institutions, high-quality institutions are believed to enhance environmental sustainability. A good institutional environment can attract more foreign direct investment, thereby promoting economic growth and technological innovation [21]. Environmental regulation is considered to affect the achievement of regional sustainable development through different mechanisms, including economic growth, social well-being, and environmental protection [22]. In terms of technology, the advancement of E-Government has a substantial positive impact on the sustainable development of countries along the Belt and Road. It promotes the digital transformation of government governance through information technology and digital tools and achieves regional sustainable development goals [23]. In terms of industry, sustainable entrepreneurship not only immediately drives economic growth but also indirectly promotes economic development via technological breakthroughs and the optimization of resource allocation [24]. Renewable energy transition has a beneficial effect on both the economy and the environment through technological innovation, policy support, and the refinement of market mechanisms [25]. The role of tourism in promoting regional sustainable development has also attracted attention [26,27].

Industrial integration, as an important trend in modern economic development, provides new pathways for regional high-quality development and efficiency improvement by breaking industry boundaries and optimizing resource allocation. Studies have shown that industrial convergence can significantly enhance regional metabolic efficiency and energy efficiency [28,29] and promote economic growth through technological innovation and industrial upgrading [30]. Agro-tourism integration, as a form of industrial convergence, may bring new development opportunities to rural areas by integrating agricultural and tourism resources, reducing transaction costs, and facilitating technological spillovers, thereby achieving a dual improvement in economic and environmental gains. Agro-tourism integration refers to a mode of industrial integration that combines the agricultural industry chain with rural tourism under the guidance of the government [31]. China has a relatively large ratio of the primary industry and abundant tourism resources, and recently, it has taken a variety of measures to promote in-depth integration between agriculture and tourism.

The pathways through which agricultural resources empower the development of rural tourism include the following: First, agricultural production scenes (such as terraced fields and tea gardens) are transformed into unique tourist landscapes, creating an “agriculture + sightseeing” appeal. Second, agricultural activities (such as picking and plowing) are extended into experiential tourism projects, realizing the value-added of “agriculture + participation” [32,33]. Third, agricultural products are developed into tourism goods through branding, forming an “agriculture + consumption” industry chain [34]. This integration provides sustainable benefits to rural communities: In terms of the economy, tourism revenue is channeled back to support agricultural upgrades and the improvement of infrastructure [35]. Socially, local farmers gain employment opportunities by running guesthouses and providing tour guide services, alleviating poverty and enhancing community cohesion [27]. Ecologically, the synergy between the protection of agricultural heritage and ecotourism reduces environmental degradation [26]. Thus, the agro-tourism integration model has the potential to become a systemic solution that links agricultural resource endowment, innovative tourism development, and multidimensional sustainable development in rural areas.

In summary, the seamless merging of agriculture and tourism can break the single-industry pillar model of agricultural regions. This model can inject new vitality into local development and has the capacity to form a sustained and virtuous cycle under reasonable operation. Most of China’s agricultural tourism resources are distributed in counties. However, previous studies did not sink the perspective to the county level. The path to improving the sustainable development of counties through agro-tourism integration is still unclear. Based on the preceding, this study introduces three research questions: 1. What is the impact of agro-tourism integration on the sustainable development of counties? 2. What are the mechanisms through which agro-tourism integration affects the sustainable development of counties? 3. Is there a regional difference in the policy effect of the agro-tourism integration? To answer these questions, this study develops hypotheses and conducts empirical tests in the following sections.

2.2. Hypotheses Development

To assess the impact of the agro-tourism integration on the sustainable development of counties and reveal the mechanism and regional differentiated effects, this study develops five research hypotheses. Hypothesis 1 explores the direct impact of the agro-tourism integration on the sustainable development of counties. Hypotheses 2 to 4 explore the mediating effects of consumer demand, capital inflow, and industrial structure. Hypothesis 5 probes the heterogeneity of policy effects across different areas.

Agro-tourism integration has multiple favorable impacts on the progress of counties. Concerning economic development, agro-tourism integration promotes the sustainable development of the agricultural economy by improving the efficiency of agricultural resource utilization [36], promotes the modernization of rural industries, and enhances the level of regional economic development. In terms of social development, it promotes agricultural technological innovation and increases non-agricultural employment demand in rural areas [31], promotes the construction of rural infrastructure and increases social welfare resources, as well as inherits rural cultural customs and promotes excellent Chinese culture [37,38]. In terms of environmental protection, agro-tourism integration improves the utilization rate of natural resources and can effectively promote ecological and environmental protection [39,40,41,42]. It appears that the development of agro-tourism integration can encourage the renovation of industrial structure in counties, optimize spatial layout, enhance economic resilience [43], provide social welfare, inherit rural culture [44], and strengthen ecological and environmental protection [45], thereby promoting the sustainable development of economy, society, and environment in counties. Drawing from the above, the forthcoming hypothesis is proposed:

H1. 

Agro-tourism integration makes a significant positive difference in the sustainable development of counties.

As a new economic model, agro-tourism integration effectively promotes the sustainable development of counties by enhancing local consumption levels and increasing regional consumer demand. It attracts tourists through activities such as nature appreciation, cultural exploration, and educational experiences, and encourages them to pay for experiences by participating in activities like fruit picking and traditional handicraft making. The green agricultural products launched in rural areas can also attract tourists who pursue a healthy lifestyle to make purchases [46]. The increase in external consumer demand can force the diversification of local goods and services, and the spillover effect generated can also improve the consumption level of local residents. In addition, the increasing demand of tourists for personalized and customized experiences prompts agricultural tourism enterprises to increase their purchases of upstream goods and basic agricultural products in order to provide richer services and products, thus promoting the consumption upgrade of the entire agricultural tourism production chain.

The increase in county consumption due to agro-tourism integration can promote the sustainable development of counties from multiple perspectives. The increase in regional consumer demand raises residents’ income, activates the development potential of local individual businesses, and enhances economic vitality [47]. In addition, the increase in consumption raises the local government’s tax revenue, providing a basis for the government to invest in local tourism projects, improve medical and educational standards, and offer more social welfare resources, as well as perfect infrastructure and supporting facilities and beautify the natural environment. The diversified consumer demand forces local enterprises to innovate in goods and services. In order to attract more tourists, local residents and merchants also proactively seek to elevate their environmental awareness and protect natural resources. Moreover, studies have found that those who love rural tourism are mostly groups with higher education levels and stronger environmental awareness. Their arrival not only supports the local economy but also brings the concept of environmental protection, helping the local area to reach a win–win situation between the economy and the ecology [46]. Building on the above, the following hypothesis is proposed:

H2. 

Agro-tourism integration promotes sustainable development of counties through the consumer demand mechanism.

Capital increment is a key driving force affecting the level of regional development. Agro-tourism integration injects vitality into local economic development, improves residents’ living conditions, protects the ecological environment, and promotes sustainable development in counties by reducing capital outflow and attracting capital inflow. For one thing, in order to carry out the growth strategy of agro-tourism integration, local governments attract more direct investment by formulating preferential tax policies for relevant industries and optimizing the entrepreneurial environment [48]. On the flip side, under the model of agro-tourism integration, rural natural and cultural resources are transformed into economic resources, and idle resources are also effectively integrated and utilized. This will increase investors’ expectations of investment returns in the local area, not only retaining local capital but also attracting the inflow of external social capital, corporate investment, and financial institution funds [49].

The increased capital can promote the sustainable development of counties in terms of multiple aspects. The increase in capital brings in funds, technology, and management experience, which promotes technological innovation and industrial upgrading and stimulates economic growth. The increase in regional capital is conducive to expanding the scale of production, enhancing production capacity, and providing financial support for infrastructure construction, which improves the conditions of transportation, communication, energy, and other infrastructure. It can furnish a material basis for the socially sustainable development of the region. Furthermore, the government can use the increased capital to enhance its environmental supervision capabilities, balance environmental protection with economic demands [50], and promote the growth of green finance to encourage funds to flow into the progress in resource-saving technologies and eco-environmental protection sectors [51]. Based on the above, the following hypothesis is proposed:

H3. 

Agro-tourism integration promotes sustainable development of counties through the capital inflow mechanism.

According to the framework of industrial integration, the integration between multiple industries will further promote the industrial structure to develop toward a higher level and a more rational direction [2]. Agro-tourism integration takes agricultural production, farmers’ life, agricultural landscape, and farming culture as tourist attractions; promotes industrial integration; and forms a new type of industrial form and business model of “promoting tourism with agriculture and revitalizing agriculture with tourism”, which promotes the comprehensive improvement of economic, social, and ecological benefits in counties. On the one hand, the model of the agro-tourism integration has given birth to a variety of tourism industry forms, such as individual businesses, pastoral complexes, and shared farms [52], prompting the rural industrial composition to shift from a single traditional agriculture to a diversified and combined type, expanding the industrial chain of tourism services, cultural creativity, and other fields and achieving the combination of the primary, secondary, and tertiary industries. In addition, the application of digital and artificial intelligence technologies has also brought about technological and management innovations, giving birth to new industrial forms such as the 3D agricultural tourism resources database in Hunan and smart agriculture in Suining, Sichuan (smart devices monitoring the growth of crops in real-time). On the other hand, agro-tourism integration enhances the added value of products and services through brand building, product design, and the expansion of marketing channels; activates the potential of industries such as transportation, catering, accommodation, and e-commerce; promotes significant growth in the service industry; and promotes the gradual transition of the industrial structure to a higher standard.

The refinement of industrial structure also means the upgrading of the regional economic structure, which brings sustainable development momentum to economic development. The diversification of industries can create more diversified employment opportunities, promote social inclusion and social equity, and prompt the continuous improvement of infrastructure and social welfare. In counties integrating agriculture and tourism, the reorganization and development of the industrial structure, led by the upgrading of the service industry, depends on the sustainable aesthetic value of natural resources and the experiential value of the ecological environment. Therefore, this transformation will promote the local improvement of the utilization effectiveness of natural resources, push traditional industries to transform into eco-friendly industries, and form a green and low-carbon development trend. Considering the above, the following hypothesis is presented:

H4. 

Agro-tourism integration promotes sustainable development of counties through the industrial structure mechanism.

Due to differences in geographical location, natural environment, historical culture, and economic strength, Chinese counties show significant regional differences, which will, to some extent, affect the contribution of the agro-tourism integration to sustainable development of counties. In terms of natural resources, differences in resource endowment and climatic conditions among counties lead to differences in the natural conditions for developing agro-tourism projects. Counties with rich natural resources can address the demands of diverse tourists and better leverage the advantages of the integration of agriculture and tourism, thereby increasing tourism consumption and promoting county economic development and sustainable development [53]. Differences in economic development levels among territories lead to differences in infrastructure and industrial foundations. Economically more developed areas have sound infrastructure and strong industrial foundations, which are conducive to agro-tourism integration and the formation of industrial cluster effects. Meanwhile, although economically less developed areas have weak infrastructure, they can better promote infrastructure construction and economic development through agro-tourism integration, resulting in a significant increase in county-level sustainable development and economic growth under the effect of policies [4]. In addition, regional differences in local policies and talents also affect the speed and effectiveness of the agro-tourism integration and thus affect its contribution to the sustainable development of counties [54]. Based on above, the ensuing hypothesis is proposed:

H5. 

Regional heterogeneity affects the contribution of the agro-tourism integration to sustainable development of counties.

2.3. Conceptual Model

The conceptual model of this study is shown in Figure 1.

3. Methodology and Data

3.1. Methodology

3.1.1. Benchmark Regression Model

To test the hypotheses proposed above, this study regards the national leisure agriculture and rural tourism demonstration county policy as a quasi-natural experiment and uses the PSM-DID method to evaluate the impact of the policy on sustainability in county regions. The benchmark model is set as follows:

$${\mathrm{S}\mathrm{D}\mathrm{C}}_{\mathrm{i}\mathrm{t}}={\mathsf{\alpha }}_0+{\mathsf{\alpha }}_1{\mathrm{A}\mathrm{T}\mathrm{I}}_{\mathrm{i}\mathrm{t}}+{\mathsf{\alpha }}_2{\mathrm{C}\mathrm{o}\mathrm{n}\mathrm{t}\mathrm{r}\mathrm{o}\mathrm{l}\mathrm{s}}_{\mathrm{i}\mathrm{t}}+{\mathsf{\gamma }}_{\mathrm{i}}+{\mathsf{\mu }}_{\mathrm{t}}+{\mathsf{\epsilon }}_{\mathrm{i}\mathrm{t}}$$

(1)

In Equation (1), the dependent variable SDC_it represents the level of sustainable development of county i in year t. The variable ATI_it indicates whether county i is a demonstration county for leisure agriculture and rural tourism in year t (including counties that were designated as demonstration counties in the current year and those designated in previous years and continued to the current year), taking a value of 1 for the treatment group samples and 0 otherwise for the control group samples. Controls_it represents a set of control variables, γ_i denotes county-level fixed effect, μ_t represents year fixed effect, and ε_it is the random disturbance term. The regression coefficient α₁ of ATI_it is the focus of the study, which measures the net effect of agro-tourism integration on improving the level of sustainable development of counties. Under the full-sample condition, if α₁ is significantly greater than 0, it indicates that the agro-tourism integration will help promote sustainable development of counties.

3.1.2. Robustness Test Method

This study conducts robustness tests by employing placebo tests, replacing the dependent variable, and isolating the impact of other policies [55,56,57].

The placebo test is conducted by randomly selecting a portion of the sample counties as a “fake” treatment group and constructing a fictitious policy variable.

The placebo test is carried out by randomly choosing a segment of the sample regions to serve as a “fake” treatment group and creating a fictitious policy variable, followed by regression analysis to check whether the research results are due to random factors. If the regression outcomes of the treated sample are insignificant, it indicates that the policy effect is genuine.

The method of replacing the dependent variable involves redefining or remeasuring the dependent variable and then conducting DID analysis again to check whether the research results are sensitive to the definition and measurement of the variable. If the policy effect remains significant after changing the definition or measurement of the dependent variable, it indicates high robustness.

Isolating the impact of other policies means incorporating other policies that might affect regional sustainablity in the observation phase into the DID model for analysis again to examine whether the impact of the ATI on the SDC changes under the impact of other policies. If the original policy remains significant after the addition of new policies, it proves that the effect of the original policy is robust.

3.1.3. Mechanism Test Model

Mechanism testing aims to identify the specific pathways through which the agro-tourism integration affects sustainable development of counties. Referring to [58,59], the model for mediating effect testing is as follows:

$${\mathrm{M}\mathrm{E}}_{\mathrm{i}\mathrm{t}}={\mathsf{\alpha }}_0+{\mathsf{\beta }}_1{\mathrm{A}\mathrm{T}\mathrm{I}}_{\mathrm{i}\mathrm{t}}+{\mathsf{\alpha }}_2{\mathrm{C}\mathrm{o}\mathrm{n}\mathrm{t}\mathrm{r}\mathrm{o}\mathrm{l}\mathrm{s}}_{\mathrm{i}\mathrm{t}}+{\mathsf{\gamma }}_{\mathrm{i}}+{\mathsf{\mu }}_{\mathrm{t}}+{\mathsf{\epsilon }}_{\mathrm{i}\mathrm{t}}$$

(2)

$${\mathrm{S}\mathrm{D}\mathrm{C}}_{\mathrm{i}\mathrm{t}}={\mathsf{\alpha }}_0+{\mathsf{\beta }}_2{\mathrm{A}\mathrm{T}\mathrm{I}}_{\mathrm{i}\mathrm{t}}+\mathsf{\theta }{\mathrm{M}\mathrm{E}}_{\mathrm{i}\mathrm{t}}+{\mathsf{\alpha }}_2{\mathrm{C}\mathrm{o}\mathrm{n}\mathrm{t}\mathrm{r}\mathrm{o}\mathrm{l}\mathrm{s}}_{\mathrm{i}\mathrm{t}}+{\mathsf{\gamma }}_{\mathrm{i}}+{\mathsf{\mu }}_{\mathrm{t}}+{\mathsf{\epsilon }}_{\mathrm{i}\mathrm{t}}$$

(3)

In Equation (2), ME_it stands for the mediating variables (consumer demand, capital inflow, and industrial structure), and the coefficient β₁ indicates the impact of ATI_it on the mediating variable. In Equation (3), the coefficient β₂ represents the direct effect of ATI_it on SDC_it after considering the impact of the mediating variable, and the coefficient θ indicates the impact of ME_it on SDC_it after controlling for the effect of ATI_it. Under the full sample condition, if β₁, β₂, and θ are all significantly greater than 0, it indicates that ME_it has a partial mediating effect.

3.1.4. Heterogeneity Analysis Method

Heterogeneity analysis refers to grouping the samples based on key characteristics and separately estimating the policy effects for each group. The purpose is to explore the differences in policy effects across different sample groups, to reveal which groups the policy is more effective for, and to provide a basis for targeted policy making. This study draws on the study by Zuo et al. (2025) and Ling et al. (2024) [60,61] and conducts regression analysis on the sample counties classified into three areas, eastern, western, and central, to compare the effectiveness of the agro-tourism integration on sustainable development of counties across different regions.

3.2. Variables

3.2.1. Dependent Variable

This study refers to studies by Xue et al. (2024), Wang et al. (2023), and Jiang et al. (2024) [10,36,62] and uses the entropy weight method to measure the level of sustainable development of counties (SDC) based on a three-level indicator system.

This study follows the classic triple bottom line theory [16] and adopts economic sustainability, social sustainability, and environmental sustainability as the primary indicators. The adoption of the secondary indicators mainly comes from the main points in the latest policy released by the Chinese central government, the “Comprehensive Rural Revitalization Plan (2024–2027) [63]” (hereinafter referred to as the rural revitalization plan), and previous studies on provincial and municipal sustainable development.

Table 1. Measurement indicators for sustainable development in counties.

Primary Indicators	Secondary Indicators	Tertiary Indicators	Attribute
Economic Development Sustainability	Resident income level	Per capita GDP (CNY)	+
	Market Economy Activity	Number of newly registered enterprises in the current year/Population	+
	Financial Market Activity	End-of-year balance of various loans from financial institutions/Population	+
	Digital Economy Level	Total number of enterprises with Weibo accounts/Population	+
Social Development Sustainability	Educational Resources	Number of full-time teachers in primary and secondary schools/Population	+
	Social Welfare Resources	Number of beds in various social welfare institutions/Population	+
	Medical Resources	Number of beds in hospitals and health centers/Population	+
	Innovation Level	Number of patents/Population	+
	Mobile Communication Level	Number of fixed-line telephone users (households)/Population	−
Environmental Development Sustainability	Carbon Emission Intensity	Carbon emissions (tons)/GDP	−
	Sulfur Dioxide Emission Intensity	Sulfur dioxide (tons)/GDP	−
	Nitrogen Oxides Emission Intensity	Nitrogen oxides (tons)/GDP	−
	PM2.5 Intensity	PM2.5 (μg/m3)/GDP	−

Note: All population units are in ten thousands, and all monetary units are in CNY ten thousand. The same applies below.

lists the secondary indicators. The selection of the tertiary indicators follows the principles of scientific soundness, reasonableness, and data accessibility and is adapted from the measurement indicators selected in previous related studies. In addition, considering that relative indicators have advantages in precision and comprehensiveness compared with absolute indicators [64], all tertiary indicators in this study are relative indicators. Below are explanations for some of the indicators:

The measurement indicator for economic sustainability: referring to the view that the number of newly registered enterprises can, to some extent, reflect the economic activity of a region [65], this study uses the ratio of the current year’s newly registered enterprises to the population to represent market economic activity.

The measurement indicator for social sustainability: referring to the view that the scale and service capacity of social welfare institutions can indicate the extent of social welfare advancement in an area [66,67,68], this study uses the total beds in assorted social welfare organizations institutions to measure the level of social welfare resources. It measures the actual capacity of non-profit social welfare institutions (such as children’s welfare homes and welfare-based nursing homes) that provide food and accommodation based on the number of beds [69].

The measurement indicators for environmental sustainability: referring to the methods used by Wang et al. (2019), Guo et al. (2022), and Wang et al. (2023) for measuring environmental sustainability with relative indicators [65,70,71], and considering the availability of data, this study selects a series of indicators (see Table 1).

The EWM, an objective weighting method, is used in this study to process panel data from 1749 Chinese counties (2008–2021). Data standardization is applied to address positive/negative impacts of indicators and eliminate measurement units. The detailed procedure for calculation is shown below.

Processing of positive and negative indicators:

$${\mathrm{y}}_{\mathrm{i}\mathrm{j}}={\displaystyle \frac{{\mathrm{x}}_{\mathrm{i}\mathrm{j}}-\mathrm{m}\mathrm{i}\mathrm{n}\left({\mathrm{x}}_{\mathrm{i}\mathrm{j}}\right)}{\mathrm{m}\mathrm{a}\mathrm{x}\left({\mathrm{x}}_{\mathrm{i}\mathrm{j}}\right)-\mathrm{m}\mathrm{i}\mathrm{n}\left({\mathrm{x}}_{\mathrm{i}\mathrm{j}}\right)}}$$

(4)

$${\mathrm{y}}_{\mathrm{i}\mathrm{j}}={\displaystyle \frac{\mathrm{m}\mathrm{a}\mathrm{x}\left({\mathrm{x}}_{\mathrm{i}\mathrm{j}}\right)-{\mathrm{x}}_{\mathrm{i}\mathrm{j}}}{\mathrm{m}\mathrm{a}\mathrm{x}\left({\mathrm{x}}_{\mathrm{i}\mathrm{j}}\right)-\mathrm{m}\mathrm{i}\mathrm{n}\left({\mathrm{x}}_{\mathrm{i}\mathrm{j}}\right)}}$$

(5)

Here, y_ij is the standardized result value of the j-th indicator x_ij for the i-th county, and max(x_ij) and min(x_ij), respectively, are maximal and minimal values of the j-th indicator among the n counties.

After standardization, calculate the proportion (P_ij) of each indicator for each county, the information entropy (I_i), and the weight (W_j). Then, compute the comprehensive evaluation index score for sustainable development (SDC_i) in each county.

$${\mathrm{P}}_{\mathrm{i}\mathrm{j}}={\displaystyle \frac{{\mathrm{y}}_{\mathrm{i}\mathrm{j}}}{{\sum }_{\mathrm{i}=1}^{\mathrm{n}}{\mathrm{y}}_{\mathrm{i}\mathrm{j}}}}, (\mathrm{j} = 1, 2, \dots , \mathrm{m})$$

(6)

$${\mathrm{I}}_{\mathrm{j}}=-{\displaystyle \frac{1}{\ln \mathrm{n}}}{\sum }_{\mathrm{i}=1}^{\mathrm{n}}{\mathrm{P}}_{\mathrm{i}\mathrm{j}}\ln {\mathrm{P}}_{\mathrm{i}\mathrm{j}}, (\mathrm{j} = 1, 2, \dots , \mathrm{m})$$

(7)

$${\mathrm{W}}_{\mathrm{j}}={\displaystyle \frac{1-{\mathrm{I}}_{\mathrm{j}}}{{\sum }_{\mathrm{j}=1}^{\mathrm{m}}1-{\mathrm{I}}_{\mathrm{j}}}}$$

(8)

$${\mathrm{S}\mathrm{D}\mathrm{C}}_{\mathrm{i}}={\sum }_{\mathrm{j}=1}^{\mathrm{m}}{\mathrm{W}}_{\mathrm{j}}{\mathrm{y}}_{\mathrm{i}\mathrm{j}}$$

(9)

3.2.2. Core Explanatory Variable

The core explanatory variable is agro-tourism integration (ATI), measured by the establishment of national leisure agriculture and rural tourism demonstration counties. Using lists from the Ministry of Agriculture and Rural Affairs, this study assigns values based on the order of establishment to create the variable ATI_it. Demonstration counties are the treatment group (ATI_it = 1), while non-demonstration counties are the control group (ATI_it = 0).

3.2.3. Mechanism Variable

Consumer demand (CD): measured by the aggregate consumer retail sales divided by the population.

Capital inflow (CI): measured by the end-of-year financial institution deposit amount divided by GDP minus the outstanding loan amount divided by GDP.

Industrial structure (IS): measured by the output value of the tertiary industry as a proportion of GDP.

3.2.4. Control Variables

To mitigate the influence of extraneous factors on the sustainable development level of counties, with reference to Kiselakova Dana et al. (2020) [18], this study includes a series of control variables, which are detailed in

Table 2. Control variables.

Control Variables	Variable Explanation	Measurement Method
Regional Area (RA)	The land area occupied by the county	Administrative regional land area (KM2)
Basic Infrastructure Level (BIL)	The development status of infrastructure in the county	Total social fixed asset investment/GDP
Level of Government Intervention (LGI)	The extent of government intervention in county development	Local general budget revenue/GDP
Yangtze River Economic Zone (YREZ)	Whether the county is supported by the development policy of the Yangtze River Economic Belt	1 if the county is within the Yangtze River Economic Belt; 0 otherwise
Epidemic Situation (ES)	Whether the development of the county is affected by the COVID-19 pandemic	1 if during the COVID-19 outbreak period; 0 otherwise

.

3.3. Data Collection

The data used for this study mainly come from the China County Statistical Yearbook and the CSMAR database. The list of national leisure agriculture and rural tourism demonstration counties comes from the Ministry of Agriculture and Rural Affairs website and local government websites. Some missing data were supplemented via linear interpolation.

This study selects the period from 2008 to 2021 as the sample interval. This interval covers all the implementation years (2010–2017) of the policy for leisure agriculture and rural tourism demonstration counties. To ensure the reliability of the data, the following treatments were applied to the sample data in this study. Counties and districts that were renamed during the sample period were merged into the same research object in this study. Due to the severe data missing in some counties, the original data were screened in this study to ensure the reliability of the data and empirical results. First, county samples with a total sum of missing values of all indicators in the selected years of the original data higher than 20% were excluded from this study. Second, county samples with missing values of a single indicator higher than 25% in the year interval and those with all data missing in a certain year were excluded from this study. After the above screening, a total of 1749 counties’ panel data were collected in this study. Among them, there were 222 demonstration counties and 1527 non-demonstration counties. The list of newly added demonstration counties each year is detailed in

Table A1. The list of newly added demonstration counties for leisure agriculture and rural tourism each year.

Year	List of New Model Counties
2010	Huairou District, Jixian County, Qian’an City, Qingxu County, Zhalantun City, Qingyuan Manchu Autonomous County, Ning’an City, Chongming County, Jurong City, Anji County, Jiashan County, Yixian County, Nanjing County, Wuyuan County, Xinyu City, Yushui District, Rongcheng City, Zhengzhou City, Huiji District, Luanchuan County, Enshi City, Longhui County, Conghua City, Yangshuo County, Jiulongpo District, Pixian County, Pujiang County, Tongzi County, Tengchong County, Tianshui City, Maiji District, Yinchuan City, Xixia District, Changji City, Jinzhou New District, Fenghua City
2011	Shexian County, Weichang County, Changzhi City, Suburban District, Eerguna City, Kuandian Manchu Autonomous County, Ji’an City, Hunchun City, TieLi City, Bin County, Nanjing City, Jiangning District, Rugao City, Tonglu County, Suichang County, Jixi County, Ningguo City, Minhou County, Zhangping City, Jinggangshan City, Yiyuan County, Yantai City, Muping District, Yanling County, Xin County, Honghu City, Zhangjiajie City, Yongding District, Yueyang City, Junshan District, Gongcheng County, Baoting County, Dazu County, Chengdu City, Wenjiang District, Wenchuan County, Qujing City, Luoping County, Xi’an City, Chang’an District, Feng County, Dunhuang City, Guide County, Helan County, Urumqi County, Qingdao City, Pingdu City
2012	Anyi County
2013	Yanqing County, Luanping County, Yuci District, Uxin Banner, Liaozhong County, Fusong County, Fengman District, Hulin County, Fengxian District, Xuyi County, Xinghua City, Shangyu City, Jiangshan City, Yingshang County, Zhangtai County, Shunchang County, Jing’an County, Shicheng County, Yinan County, Daiyue District, Queshan County, Gucheng County, Guiyang County, Xinxing County, Lingchuan County, Qianjiang District, Cangxi County, Pingchang County, Leishan County, Xingyi City, Yulong Naxi Autonomous County, Mile City, Pingli County, Yongjing County, Huangzhong County, Wuzhong City, Litong District, Bohu County, Xinjiang Production and Construction Corps Wujiaqu City
2014	Pinggu District, Yuanshi County, Chengde City, Shuangluan District, Yangcheng County, Chifeng City, Keshiketeng Banner, Benxi Manchu Autonomous County, Changchun City, Shuangyang District, Mulan County, Taizhou City, Jiangyan District, Yixing City, Lanxi City, Xinchang County, Huoshan County, Taining County, Liancheng County, Wuning County, Sishui County, Linqu County, Dengfeng City, Yuan’an County, Xinhua County, Mayang Miao Autonomous County, Boluo County, Longsheng Autonomous County, Qionghai City, Wulong County, Wusheng County, Fenggang County, Chengjiang County, Zhaoshui County, Liangdang County, Menyuan Hui Autonomous County, Jinfeng District, Manas County, Zhuanghe City, Ninghai County, Xinjiang Production and Construction Corps 10th Division 185th Regiment Beiting City
2015	Daxing District, Wuqing District, Lincheng County, Tangshan City, Funan District, Shuangqiao District, Pingshun County, Taigu County, Hulunbuir City, Arong Banner, Dawo County, Gaizhou City, Hui’an County, Jiaohe City, Harbin City, Acheng District, Muleng City, Dafeng City, Haimen City, Shuyang County, Lishui District, Tiantai County, Kaihua County, Deqing County, Huangshan City, Huangshan District, Jing County, Songxi County, Nanchang County, Shangyou County, Fuliang County, Qingzhou City, Qufu City, Zaozhuang City, Shanting District, Shangcheng County, Mengjin County, Fengqiu County, Suiping County, Yingshan County, Nanzhang County, Xianfeng County, Liuyang City, Chenzhou City, Beihu District, Leiyang City, Dapu County, Nanxiong City, Mengshan County, Luchuan County, Ding’an County, Tongliang District, Wansheng Economic and Technological Development Zone, Kaixian County, Luzhou City, Naxi District, Jiangyou City, Xichong County, Ya’an City, Anshun City, Xixiu District, Jiangkou County, Luxi County, Yanjin County, Jiangzi County, Nedong County, Liuba County, Hezheng County, Haidong City, Ledu District, Pingluo County, Zepu County, Zhaosu County, Lvshunkou District, Laoshan District, Xiangshan County, Xinjiang Production and Construction Corps 8th Division 150th Regiment
2016	Changping District, Fangshan District, Baodi District, Pingshan County, Guantao County, Zhangjiakou City, Chongli District, Taiyuan City, Jiancaoping District, Quwo County, Wuyuan County, Ningcheng County, Xinbin Manchu Autonomous County, Suizhong County, Jiutai District, Antu County, Dongning City, Lindian County, Qidong City, Changzhou City, Jintan District, Pukou District, Yancheng City, Yandu District, Lishui City, Tongxiang City, Pujiang County, Nanling County, Jinzhai County, Lingbi County, Yongtai County, Ji’an City, Qingyuan District, Yichun City, Yuanzhou District, Gan County, Lanning County, Penglai City, Luoyang City, Minquan County, Guangshan County, Yichang City, Yiling District, Enshi Tujia and Miao Autonomous Prefecture, Anren County, Jishou City, Pingyuan County, Wengyuan County, Rong County, Guigang City, Tantang District, Lipu County, Hechuan District, Tongnan District, Xichang City, Langzhong City, Taijiang County, Libo County, Bijie City, Bailingdujuan Management District, Guangnan County, Shuifu County, Jiacha County, Mangkang County, Suo County, Xianyang City, Weicheng District, Lantian County, Weinan City, Linwei District, Dali County, Yangling Demonstration Zone, Kang County, Zhuanglang County, Qilian County, Huzhu County, Pengyang County, Fukang City, Hejing County, Chabuchar Sibe Autonomous County, Wafangdian City, Huangdao District, Xinjiang Production and Construction Corps 1st Division 10th Regiment
2017	Xingtai County, Xinglong County, Chengde County, Rui Cheng County, Yijinhuoluo Banner, Donggang City, Yingkou City, Bayuquan District, Tonghua County, Wangqing County, Hailin City, Wuchang City, Xuzhou City, Jiawang District, Suqian City, Suyu District, Dongtai City, Yongjia County, Quzhou City, Kecheng District, Liandu District, Xiuning County, Qianshan County, Shouning County, Youxi County, Fuqing City, Yongxiu County, Nanfeng County, Chongyi County, Wulian County, Zhucheng City, Boai County, Lushi County, Wuhan City, Dongxihu District, Daye City, Guxiang County, Pingjiang County, Guangzhou City, Zengcheng District, Zhuhai City, Doumen District, Mashan County, Beiliu City, Tiandong County, Fuling District, Qijiang District, Luojiang County, Gao County, Suining City, Chuanshan District, Guiding County, Jianshui County, Lijiang City, Gucheng District, Basu County, Shiquan County, Huayin City, Taibai County, Qinzhou District, Wuwei City, Liangzhou District, Huangyuan County, Longde County, Zhongwei City, Shapotou District, Shawan County, Changhai County, Jimo City, Beilun District, Xinjiang Production and Construction Corps 1st Division 7th Regiment

in Appendix A.

4. Empirical Results

4.1. PSM Results

This study uses all control variables as covariates and employs logit regression to estimate the propensity scores of the samples. Nearest neighbor matching is applied within the common support domain to match the samples, with the matching scores shown in Figure 2. Before matching, there were 24,304 samples, and after matching, 23,437 samples remained for further analysis. The results of the PSM balance test are presented in

Table 3. The results of the PSM balance test.

Variables	U or M	Mean		Bias	Reduct Bias	t-Test
		Treated	Control			T	p > t
RA	U	7.6304	7.2754	42.4		15.20	0.000
	M	7.6281	7.6411	−1.5	96.3	−0.51	0.613
BIL	U	1.0682	0.87453	29.7		12.44	0.000
	M	1.0649	1.0657	−0.1	99.6	−0.03	0.972
LGI	U	0.734	0.06833	7.1		2.63	0.009
	M	0.7342	0.07728	−5.4	23.9	−1.40	0.160
YREZ	U	0.44193	0.41471	5.5		2.21	0.027
	M	0.44244	0.43953	0.6	89.3	0.17	0.864
ES	U	0.25784	0.13409	31.6		14.20	0.000
	M	0.25698	0.2404	4.2	86.8	1.10	0.270

Note: U = unmatched, M = matched.

, indicating that there are no significant differences between the treatment group and the control group across all control variables. This suggests that the PSM method is effective in this study, providing a balanced sample basis for the subsequent causal effect analysis.

4.2. DID Results

4.2.1. Parallel Trend Test

Before conducting the DID analysis, it is necessary to conduct a parallel trend test for the explanatory variables, which is a prerequisite for the subsequent analysis. The results of the parallel trend test are shown in Figure 3 and Figure 4. Figure 3 shows that the trends of the treatment group and the control group were parallel until 2010, when the policy was implemented, and the trends changed significantly after 2010. Figure 4 describes the policy treatment effects at different policy implementation time points. Before the policy implementation, the estimated coefficients were basically not statistically significant (the confidence intervals included 0), while after the policy implementation, the estimated coefficients were significant (the confidence intervals did not include 0). This means that before the policy implementation, there was no significant difference in the level of sustainable development between the counties in the treatment group and the control group, which is in line with the parallel trend assumption. The above analysis shows that the parallel trend test was passed in this study, and the use of the PSM-DID method is appropriate.

4.2.2. Benchmark Regression

This study proceeded by conducting a DID analysis after performing a PSM analysis.

Table 4. Benchmark regression results.

	(1)	(2)	(3)	(4)	(5)
	SDC	SDC	SDC	SDC	SDC
ATI	0.003 ***	0.003 ***	0.003 ***	0.003 ***	0.003 ***
	(3.67)	(3.74)	(3.73)	(3.76)	(3.76)
RA	0.003	0.003	0.003	0.003	0.003
	(1.43)	(1.35)	(1.34)	(1.33)	(1.33)
BIL		−0.002 ***	−0.002 ***	−0.002 ***	−0.002 ***
		(−6.71)	(−6.78)	(−6.80)	(−6.80)
LGI			0.006	0.006	0.006
			(1.39)	(1.41)	(1.41)
YREZ				−0.003 *	−0.003 *
				(−1.86)	(−1.86)
ES					0.014 ***
					(35.09)
_cons	−0.009	−0.007	−0.007	−0.005	−0.005
	(−0.63)	(−0.47)	(−0.47)	(−0.39)	(−0.39)
County-level fixed effect	Yes	Yes	Yes	Yes	Yes
Year fixed effect	Yes	Yes	Yes	Yes	Yes
N	23,437	23,437	23,437	23,437	23,437
adj. R2	0.335	0.340	0.340	0.341	0.341

Note: t statistics in parentheses, * p < 0.1, ** p < 0.05, and *** p < 0.01, the same below.

reports the benchmark regression results of the impact of the demonstration county policy for leisure agriculture and rural tourism on the sustainable development of counties. This study adopts a progressive regression strategy, gradually adding a series of control variables from column (1) to column (5). That is, each column adds one control variable on the basis of the previous column for regression analysis. The results show that the ATI remains significantly positive at the 1% significance level, even after the inclusion of other control variables. The results show that the policy for demonstration counties in leisure agriculture and rural tourism has a significant driving effect on the sustainable development of counties. H1 is supported, answering research question 1: “What is the impact of agro-tourism integration on sustainable development of counties?”.

4.3. Robustness Test

4.3.1. Placebo Test

In this study, we randomly selected the same number of counties from the research sample each year according to the number of demonstration counties for leisure agriculture and rural tourism, generated a set of random numbers, constructed a “fake” policy dummy variable (Random-did), and used it as a substitute for ATI to conduct 500 repeated simulation regressions with the explained variable (SDC), obtaining the kernel density distribution of 500 Random-did estimated coefficients. The placebo test results are shown in Figure 5, where the regression coefficients obtained from the random sampling are distributed on both sides of zero. However, in the real benchmark regression, the coefficient of the interaction term for the demonstration county policy (ATI) is 0.003, which is greater than the vast majority of simulated values and is a clear outlier. It is a low-probability event. Therefore, the sustainable development effect of the development of agro-tourism integration in counties is not caused by routine random factors and unobservable factors. The benchmark regression results are robust to some extent.

4.3.2. Replacement of the Dependent Variable

There are various ways to measure the sustainable development of counties. This study refers to the studies of Zhang (2024) and Alberto et al. (2025) [72,73], and uses per capita GDP (Per_GDP) as the alternative dependent variable for retesting with the DID model. As shown in column (1) of Table 5, after the dependent variable is replaced with Per_GDP data, the core explanatory variable ATI remains significantly positive at the 5% level. That is, after measuring the level of county sustainable development in different ways, the promoting effect of agro-tourism integration on the sustainable development of counties still remains significant. This result once again confirms that the core conclusion is robust and reliable.

4.3.3. Isolation from Other Policy Interferences

In the actual operation of real-world economic and social systems, many similar or related policies are implemented simultaneously or in an overlapping manner across regions. The implementation of other policies may introduce biases in the assessment of the effects of the leisure agriculture and rural tourism demonstration county policy. Based on a review of existing literature [74], we found that during the sample period, the pilot policies to support migrant workers and others in returning to their hometowns to start businesses could potentially affect the sustainable development level of local counties. To exclude the impact of this policy, this study includes the pilot policy (PP) as a dummy variable in the benchmark regression control variables, with counties included in the pilot for returning to start businesses taking a value of 1, and others taking a value of 0. The regression results are shown in column (2) of Table 5. The results show that after considering the interference of other policies, the coefficient of the core explanatory variable ATI remains significantly positive at the 1% level. This indicates that other policy shocks have not affected the promoting effect of agro-tourism integration on county sustainable development. The robustness tests conducted above further enhance the persuasiveness of the benchmark regression results.

Table 5. Robustness test results.

	(1)	(2)
	Per_GDP	SDC
ATI	0.052 *	0.003 ***
	(1.96)	(3.70)
RA	0.101 **	0.002
	(2.02)	(1.29)
BIL	−0.042 ***	−0.002 ***
	(−2.83)	(−6.86)
LGI	−0.596 ***	0.006
	(−4.38)	(1.49)
YREZ	−0.095 *	−0.003 *
	(−1.82)	(−1.75)
ES	1.298 ***	0.014 ***
	(78.89)	(33.96)
PP		0.001 **
		(2.18)
_cons	9.095 ***	−0.005
	(24.59)	(−0.34)
County-level fixed effect	Yes	Yes
Year fixed effect	Yes	Yes
N	23,437	23,437
adj. R2	0.772	0.339

Table 5. Robustness test results.

	(1)	(2)
	Per_GDP	SDC
ATI	0.052 *	0.003 ***
	(1.96)	(3.70)
RA	0.101 **	0.002
	(2.02)	(1.29)
BIL	−0.042 ***	−0.002 ***
	(−2.83)	(−6.86)
LGI	−0.596 ***	0.006
	(−4.38)	(1.49)
YREZ	−0.095 *	−0.003 *
	(−1.82)	(−1.75)
ES	1.298 ***	0.014 ***
	(78.89)	(33.96)
PP		0.001 **
		(2.18)
_cons	9.095 ***	−0.005
	(24.59)	(−0.34)
County-level fixed effect	Yes	Yes
Year fixed effect	Yes	Yes
N	23,437	23,437
adj. R2	0.772	0.339

It can be seen from this that Hypothesis 1 passed a series of robustness tests and was sufficiently supported.

4.4. Mechanism Test

This study employs a mediation effect model to examine the mechanisms of consumer demand, capital inflow, and industrial structure. The results are presented in

Table 6. Results of the mediation effect mechanism test.

	(1)	(2)	(3)	(4)	(5)	(6)
	CD	SDC	CI	SDC	IS	SDC
ATI	0.057 **	0.003 ***	0.144 ***	0.003 ***	0.009 *	0.003 ***
	(2.08)	(3.79)	(3.77)	(3.82)	(1.95)	(3.73)
RA	0.148	0.002	0.133	0.002	0.012	0.002
	(1.21)	(1.16)	(0.88)	(1.18)	(0.53)	(1.28)
BIL	−0.051 ***	−0.002 ***	−0.088 ***	−0.002 ***	0.009 ***	−0.002 ***
	(−3.19)	(−6.16)	(−5.24)	(−6.08)	(3.21)	(−6.86)
LGI	−0.835 ***	0.008 *	0.014	0.006	0.126 ***	0.005
	(−5.04)	(1.83)	(0.04)	(1.41)	(3.94)	(1.14)
YREZ	−0.105	−0.002*	−0.054	−0.003 *	−0.022 **	−0.002 *
	(−1.52)	(−1.74)	(−0.94)	(−1.82)	(−2.11)	(−1.69)
ES	1.421 ***	0.011 ***	−0.357 ***	0.015 ***	0.122 ***	0.013 ***
	(77.49)	(7.23)	(−20.92)	(33.53)	(37.33)	(20.59)
CD		0.002 **
		(2.33)
CI				0.001 *
				(1.88)
IS						0.009 **
						(2.23)
_cons	7.607 ***	−0.023	−1.266	−0.004	0.246	−0.008
	(8.37)	(−1.51)	(−1.13)	(−0.25)	(1.49)	(−0.55)
County-level fixed effect	Yes	Yes	Yes	Yes	Yes	Yes
Year fixed effect	Yes	Yes	Yes	Yes	Yes	Yes
N	23,437	23,437	23,437	23,437	23,437	23,437
adj. R2	0.755	0.346	0.212	0.342	0.524	0.344

. Columns (1) and (2) present the results of the mediating effect test of consumer demand (CD). In column (1), after controlling for other influencing factors and the double fixed effects of time and county, agro-tourism integration (ATI) is significantly positive at the 5% level, indicating a significant positive correlation between ATI and CD. This suggests that the integration of agriculture and tourism is conducive to promoting the growth of consumer demand. In column (2), both ATI and CD have a significant impact on county sustainable development (SDC) at the 1% level, indicating that CD has a partial mediating effect in the influence of ATI on SDC. Similarly, columns (3) and (4) present the results of the mediating effect test of capital inflow (CI), while columns (5) and (6) present the results of the mediating effect test of industrial structure (IS). The results show that consumer demand, capital inflows, and industrial structure all have partial mediating effects in the impact mechanism of agro-tourism integration on the sustainable development of counties. H2, H3, and H4 are all supported, answering research question 2: “What are the mechanisms through which agro-tourism integration affects sustainable development of counties?”.

4.5. Heterogeneity Analysis

This study, based on geographical location and taking into account factors such as economic development levels, policy orientations, and administrative divisions, divides counties across the country into three regions: the eastern region (counties belonging to Beijing, Hebei Province, Zhejiang Province, Shandong Province, Hainan Province, etc.), the central region (counties belonging to Shanxi Province, Anhui Province, Henan Province, Hubei Province, Jiangxi Province, etc.), and the western region (counties belonging to Inner Mongolia Autonomous Region, Sichuan Province, Chongqing Municipality, Gansu Province, Qinghai Province, etc.).

Following the three regions mentioned above, this study once again employs the multi-period DID method to estimate Equation (1). As can be seen in the model estimation results in

Table 7. Heterogeneity analysis results.

	(1)	(2)	(3)
	Eastern Region	Central Region	Western Region
	SDC	SDC	SDC
ATI	0.002 **	0.001 *	0.005 ***
	(2.14)	(1.66)	(2.64)
RA	−0.000	−0.002	0.004
	(−0.09)	(−0.78)	(1.46)
BIL	−0.001 **	−0.001 ***	−0.003 ***
	(−2.08)	(−4.63)	(−4.32)
LGI	−0.006	0.004	0.003
	(−1.06)	(0.78)	(0.36)
YREZ	−0.001	−0.003	−0.004
	(−0.94)	(−1.36)	(−1.17)
ES	0.015 ***	0.011 ***	0.020 ***
	(15.43)	(27.67)	(20.94)
_cons	0.012	0.024	−0.010
	(0.50)	(1.41)	(−0.51)
County-level fixed effect	Yes	Yes	Yes
Year fixed effect	Yes	Yes	Yes
N	5177	11,146	7114
adj. R2	0.325	0.410	0.332

, the coefficient of the core explanatory variable, ATI, is significantly positive in all regions, including the eastern region as well as the central and western regions, and it is significant at the 5%, 10%, and 1% levels, respectively. This indicates that the policy of leisure agriculture and rural tourism demonstration counties does not alter its promoting effect on the sustainable development level of counties due to geographical location differences, further verifying the robustness of the previous estimation results. Looking at the estimated coefficients, there are significant regional differences in the promoting effect of agro-tourism integration on the sustainable development of counties, specifically showing a stepwise decreasing pattern of “west > east > central”. Hypothesis 5 is supported, answering research question 3: “Is there a regional difference in the policy effect of the agro-tourism integration?”.

5. Discussion

5.1. Discussion of Direct Effect

As shown in Table 4 and Table 5, H1 is supported, and agro-tourism integration can significantly promote county-level sustainable development. This result revalidates the efficacy of the theory of industrial integration from the viewpoint of the integration of primary and tertiary industries and responds to the view put forward by Wang et al. (2024) that the integration of countryside industries is an effective strategy to enrich husbandmen and achieve comprehensive rural revitalization [2]. Compared with previous studies, in terms of research scope, the empirical outcomes of this study expand upon the views of Ammirato et al. (2020) regarding the promotion of economic development through agro-tourism integration, extending the policy spillover effects to social development and environmental protection [46]. This corresponds to the view of He et al. (2022) that well-constructed policies bring about comprehensive spillover effects [75]. In relation to research depth, this study broke through the research boundaries of regional sustainable development at the provincial and municipal levels, as established by Gao et al. (2021) and Sun and Wang (2022) [9,20]. It shifts the focus to the county level, responding to the call by Heather et al. (2021) for more in-depth research [76].

5.2. Discussion of Mediating Mechanism Effects

As shown in Table 6, Hypotheses 2, 3, and 4 were all supported. This indicates that agro-tourism integration can promote the sustainable development of counties by increasing consumer demand, attracting capital inflows, and optimizing the industrial structure.

Firstly, the conclusions correspond to the view of promoting the development of rural areas by boosting consumption [47] and verify the effectiveness of the policy orientation of “stimulating consumption and promoting internal circulation” for sustainable development [77]. This provides additional theoretical support for the views proposed by Danupon et al. (2025) that rural tourism scenarios should try to extend the time of tourist consumption, increase the convenience of consumption, and raise the amount of tourist spending [5].

Secondly, the results once again emphasize the importance of capital for regional development [50] and support the view that capital is a prerequisite for development [78]. However, as Tang et al. (2025) emphasized, the sustainable development of poor areas cannot rely solely on government investment; more importantly, it is necessary to boost the region’s own investment attractiveness [49]. The outcomes of this study confirm the potential worth of the agro-tourism integration in activating and uncovering the investment attractiveness of counties.

Finally, the insights gained from this study extend the conclusion of Rong et al. (2023) that industrial structure upgrading is the key to breaking the environmental–economic dilemma in poor areas to all counties nationwide [79]. As mentioned by Zhang et al. (2022), adjusting the industrial structure reasonably according to regional resource endowment is a prerequisite for sustainable development [80]. The findings of this study mirror the fact that agro-tourism integration, based on the rich natural resources of rural areas, makes their economic and social values explicit and transforms natural resources into “clean energy” for the sustainable development of counties, achieving the transfer of resource value from the agricultural sector to the service sector.

5.3. Discussion of Heterogeneity Analysis Results

The heterogeneity analysis findings point to the fact that Hypothesis 5 is verified. Regional diversity affects the contribution of the agro-tourism integration to the sustainable development of counties, with the sequence being the western region > eastern region > central region. This result indirectly verifies the view of Xiao et al. (2025) that different spatial patterns affect the level of rural tourism development [53]. Specifically, in contrast to the eastern and central regions, the western regions have superior natural resource endowments, better ecological environments, and richer historical culture and folk customs, as well as more abundant agricultural resources. However, the general economic level of the counties in the western region is comparatively low, with a single industrial structure, poor infrastructure conditions, and a smaller proportion of commercial development [81]. Therefore, the facilitating influence of agro-tourism integration on the sustainable development of counties in the western region is more significant. For example, Jiuzhaigou and Yading in Sichuan Province, which is located in the western region, have world-class natural resources and rich folk and agricultural cultures. However, due to geographical location and topographical constraints, the commercialization level of the region is low, and the economy is relatively underdeveloped. The implementation of agro-tourism integration policies can play a “timely and crucial” role, fully mobilizing the local resource endowments and achieving significant breakthroughs in the multidimensional development of the region.

Compared with the western regions, some counties in the eastern regions have unique coastal characteristic agricultural resources and coastal natural landscapes, and the majority of counties have a high original level of urbanization and industrial structure diversification [82,83]. For example, counties in Zhejiang Province have developed county economies, complete infrastructure, more mature service industries, and stronger government governance capabilities. The agro-tourism integration policy has good implementation conditions in these counties and can quickly generate value that is visible, playing an “icing on the cake” role in county development. However, since the development paths of counties in the eastern regions are diversified and agro-tourism integration is just one of them, the improvement brought by agro-tourism integration to local development is not as prominent as in the western regions.

Finally, counties in the central region have weaker natural resources compared to the western region and less developed infrastructure and service industry foundations compared to the eastern region, which results in lower tourism appeal than the western and eastern regions. For example, counties in Shanxi Province have abundant agricultural resources, but most of them are in the stage of accelerated industrialization and urbanization [84], which causes significant damage to the local natural environment and reduces the local agro-tourism appeal. Therefore, agro-tourism integration appears to have a limited effect on the development of counties in the central region.

6. Policy Implications and Conclusions

The above results answer the three research questions posed in this study and achieve the objective of exploring the impact of the agro-tourism integration on the sustainable development of counties, revealing the underlying mechanisms and regional heterogeneity effects. Furthermore, to achieve the objective of providing a scientific basis for the formulation of rural revitalization policies, this study offers the following policy implications.

Firstly, multiple measures should be adopted to increase agricultural tourism consumption. First, consumption scenarios should be enriched by introducing a variety of “experience + consumption” scenarios, adding situational retail (such as mobile stalls within scenic areas), and extending the time for tourist consumption (for example, night markets featuring local characteristics). Second, payment convenience should be improved by collaborating with Alipay/WeChat to issue special consumption coupons and launch an “agro-tourism white note” that allows tourists to pay later (with the government providing bad debt risk guarantees). Finally, a county-level consumption data platform can be established to monitor the conversion rate of consumption at each node in real time, optimize weak links in a targeted manner, and provide tax relief for innovative business models to encourage product iteration.

Second, government–enterprise cooperation should be promoted, and long-term funds should be introduced. On the part of the local government, agricultural financial assistance should be increased, and funds should be allocated to improve the service facilities of rural tourism and manage the rural landscape. It is important to make full use of idle land resources to provide land support for long-term projects of agro-tourism integration. On the part of social capital, it is necessary to attract enterprises to settle in through investment-attracting channels, especially by introducing leading enterprises to play a demonstrative role and create characteristic projects. Well-known domestic and international companies should be attracted for long-term investment while retaining local capital to sustain the long-term progress of agricultural leisure and rural travel.

Finally, the service industry should be developed in a location-specific manner and continuously optimized for the industrial structure. Counties in the eastern region that are relatively developed can build industrial clusters and plan the construction of integrated agricultural-tourism parks to leverage agglomeration effects and promote the high-quality development of agricultural tourism in their jurisdictions. Counties in the central region that focus on industrial development should extend their industrial chains, develop elaborate processing of crops and livestock products, and diversify services to boost the incremental value of farm produce. Counties in the western region, which have superior natural resources but weaker economic development, should improve service quality; strengthen infrastructure construction; and upgrade facilities such as rural roads, parking lots, and charging stations. Moreover, they should accelerate the application of digital technologies to provide virtual reality experiences and online booking platforms, thereby enhancing the convenience of consumption through smart agricultural tourism.

This study still has some limitations. First, there is still a small portion of data missing. There are numerous counties in China, and some remote counties suffer from missing statistical data. This gives rise to a particular degree of lack of generalizability in the conclusions of this study. Second, the data of this analysis are bounded by the statistical yearbook. The latest data of this study are up to 2021, lacking a statistical description and discussion of the latest development situation. Third, the perspective is limited. The internal driving mechanism of agro-tourism’s integration into the sustainable development of counties is very complex. This study only discusses it from three aspects, which is not comprehensive enough.

In brief, this study verifies the driving effect of the agro-tourism integration on the sustainable development of counties; reveals the mediating mechanisms of consumer demand, capital inflow, and industrial structure; and clarifies that the influence of the policy is more significant in the western area, providing empirical evidence for regionally differentiated policy design under the rural revitalization strategy. Future research could combine the trends of the digital economy and smart agriculture to explore new paradigms for agro-tourism’s integration with emerging technologies to make up for the limitations of data and perspectives and promote the dynamic adaptation of policy tools.