Towards sustainable agricultural systems: a TOE-based configurational analysis of sports-tourism-agriculture integration

Abstract

The integration of sports, tourism, and agriculture is a vital pathway for diversified, sustainable rural economic development and rural revitalization. To explore the multi-condition synergistic mechanisms behind the coordinated development efficiency of the three sectors within sustainable agricultural systems, this study adopts the Technology-Organization-Environment (TOE) framework. Combining necessary condition analysis (NCA) and fuzzy-set qualitative comparative analysis (fsQCA), we use 2019–2023 conditional panel data and 2020–2024 outcome data across 20 Chinese provinces (100 province-year observations) with a one-year lag for empirical analysis. The results show that digital technology innovation (d = 0.305, p = 0.005) and resource endowments (d = 0.329, p = 0.040) act as significant necessary-but-insufficient bottleneck constraints, requiring minimum thresholds to achieve high coordination efficiency. Specifically, reaching 100% efficiency demands at least 81.8% for digital innovation and 77.2% for resource endowments. No single condition alone is sufficient to generate the outcome, as all necessity test consistency values are below 0.9. This study further identifies three equivalent pathways driving sustainable integration, with overall solution consistency of 0.813 and coverage of 0.570: (a) the resource endowment–driven path, which offsets weak policy and market conditions via superior natural and cultural resources (unique coverage = 0.101); (b) the technology–market synergy path, where digital innovation paired with market demand substitutes for a solid regional economic foundation (unique coverage = 0.021); and (c) the digital innovation–resource compensation path, where technological empowerment makes up for human capital scarcity (unique coverage = 0.058). Further analysis reveals distinct regional adaptability among these pathways: China’s eastern regions primarily rely on dual drivers of technology and market demand (62.5% follow the technology–market synergy path), while central and western regions of China depend more on resource endowments and emerging digital capabilities, with 80% of central provinces and 100% of western provinces following the resource endowment–driven path (Fisher’s exact p < 0.001, Cramér’s V = 0.662). These findings validate the TOE framework’s applicability in cross-industry integration and offer differentiated policy insights to advance sustainable agricultural systems and rural economic resilience across China’s regions.

1 Introduction

Developing sustainable agricultural systems and advancing the comprehensive revitalization of rural areas are core strategies for addressing modern rural economic challenges. In this context, promoting the integration of the sports industry with the tourism and agricultural industries (referred to as “sports-tourism-agriculture” integration) has emerged as an important approach to respond to the global trend of industrial convergence and achieve diversified and sustainable rural economic development. In China, the deep integration of these sectors is of great significance for accelerating agricultural modernization, meeting diverse public needs, and implementing the “Healthy China 2030” strategy (Xu et al., 2025). Sports events serve as a powerful economic lever to stimulate consumption, integrating activities such as “sports events in scenic spots, in rural areas, and outdoors” promotes a harmonious interaction between humans and natural ecosystems, becoming an important force in driving sustainable consumption upgrading in China (Xu et al., 2025). Furthermore, steadily advancing rural revitalization in China requires prioritizing the resolution of the “three rural issues” at the core to promote the integrated development of urban and rural areas—a strategic deployment based on China’s unique agricultural conditions (Qiu, 2025). The integration of sports, tourism, and agriculture responds precisely to this strategic call. Using rural areas as spatial carriers in China, this cross-sectoral model takes agricultural resources as its foundation, endows them with sports functions, and utilizes tourism experiences as connecting links. This integration breaks down the boundaries between traditional agriculture, sports, and tourism, building a vital bridge for the flow of urban and rural elements to achieve structural reorganization and functional complementarity within the Chinese context. Against this macro-policy and theoretical background, the integration of sports, tourism, and agriculture exhibits distinct China-specific scenarios and localized evolutionary characteristics (Wu and Liu, 2022). Notably, the sustainable agricultural systems on which this integration depends are themselves shaped by complex spatiotemporal dynamics, as shown in recent research on terrestrial ecosystem productivity (Li et al., 2025) and high-quality agricultural development in representative regions of China (Xiang et al., 2024). These findings underscore that rural sustainability is inherently multidimensional, reinforcing the rationale for cross-sectoral integration approaches such as the one examined in this study in the Chinese rural context.

The collaborative concept of “sports as the stage, tourism as the performance, and agriculture as the backdrop” was proposed by scholars dates back to the 1990s. With the explosive growth of outdoor sports, leisure agriculture, and rural tourism, academic research on this integration has rapidly gained momentum, primarily focusing on general industrial integration(Benner and Ranganathan, 2013; Kim et al., 2015) and the “two-to-two” bilateral integration of these sectors(Wang et al., 2022; Dorocki et al., 2016; Fleischer and Tchetchik, 2005). However, a significant research gap remains: we lack theoretical explanation and empirical verification regarding how to systematically integrate the kinetic energy of sports, the flow of tourism, and local agricultural resources into a deep tripartite integration in the Chinese context, The complex driving mechanisms behind efficient development also remain underexplored. Moreover, existing studies on influencing factors are mostly limited to case descriptions or single-dimensional testing, lacking systematic quantitative explanations of the combined effects of multiple antecedent conditions under a comprehensive “technology-organization-environment” (TOE) framework. Endowed with the mission of boosting rural revitalization and building sustainable agricultural and sports systems in China, this integration model provides a rich and unique experiential field for exploring the aforementioned theoretical gaps.

To address these gaps, this study adopts the TOE framework and focuses on the innovative allocation of production factors to enhance agricultural sustainability. Utilizing balanced panel data covering conditions from 2019 to 2023 and outcomes from 2020 to 2024 (from 20 Chinese provinces, 100 province-year observations), employing a one-year lag design, this paper employs Necessary Condition Analysis (NCA) and fuzzy set Qualitative Comparative Analysis (fsQCA) to systematically analyze the complex formation mechanisms driving high-quality coordinated development efficiency within China’s rural contexts. The study aims to answer three core questions: First, what are the necessary conditions for improving the coordinated development efficiency of sports, tourism, and agriculture in China? Second, which conditions are more decisive in the process of improving efficiency? Third, what equivalent configurations can promote a leap in coordinated development efficiency in a “different paths but same destination” manner?

The marginal contributions of this paper are threefold. First, it systematically explores configurational pathways from the technological dimension (digital technology innovation), the organizational dimension (government policy support and cultural-tourism human resources), and the environmental dimension (market demand and resource endowment), providing a new empirical approach to the study of sports-tourism-agriculture integration in China. Secondly, by combining NCA with fsQCA in a hybrid analytical design, the study not only identifies necessary bottleneck conditions but also reveals sufficient configurational paths, achieving methodological triangulation that enhances the robustness of causal inferences. Thirdly, through regional compatibility analysis, it accurately captures the differentiated pathway–region alignment patterns across China’s eastern, central, and western macro-regions, providing significant practical insights for advancing sustainable agricultural systems and context-sensitive industrial synergy in China.

2 Theoretical foundation and variable system construction

2.1 Variable system design

2.1.1 Result variable: coordinated development efficiency of sports, tourism and agriculture

The “coordinated development efficiency” fundamentally measures the relative performance of the sports, tourism, and agricultural systems in converting input factors into “integrated outputs” within the same spatial framework. Its theoretical foundation stems from three core principles: ① Industrial Integration Theory, which highlights the “1 + 1 + 1 > 3” synergistic surplus resulting from cross-sectoral reorganization of heterogeneous resources (Benner and Ranganathan, 2013); ② The efficiency analysis paradigm (Farrell, 1957), which uses an input–output distance function to delineate the optimal boundary for maximizing integrated output with minimal input; ③ The logic of China’s “Three Transformations” rural reform, which transforms ecological resources, agricultural landscapes, and cultural heritage into productive assets within a market-oriented framework.

Conventional efficiency evaluation in industrial integration research has predominantly relied on standard self-evaluation Data Envelopment Analysis (DEA) models, such as the radial CCR model (Charnes et al., 1978) and the BCC model (Banker et al., 1984). While widely applied, these standard approaches exhibit two fundamental limitations when used complex multi-sectoral integration in Chinese provinces. First, they rely strictly on a “self-evaluation” mechanism, which allows each decision-making unit (DMU) to assign the most favorable, sometimes extreme, weights to its own inputs and outputs. This often results in multiple DMUs tied at the maximum efficiency score of 1.0, failing to provide full discrimination and ranking across provincial economies in China. Second, standard DEA assumes that DMUs operate as isolated, independent competitors. However, the integration of sports, tourism, and agriculture operates on a symbiotic, cooperative logic of “promoting tourism through sports, driving agriculture via tourism, and boosting sports through agriculture,” where inter-provincial cooperation, shared tourist flows, and resource co-recommendation are highly encouraged within China’s rural development practice.

To address these limitations, this study adopts the DEA Benevolent Cross-Efficiency Model (Doyle and Green, 1994). This model introduces a “peer-evaluation” mechanism alongside self-evaluation. Specifically, the benevolent paradigm assumes that while all DMUs seek to maximize their own efficiency, they also assign the most lenient and cooperative weight combinations to evaluate their peers. This approach mitigates the issue of extreme weight flexibility, eliminates the tie-score problem (allowing scores to be fully ranked), and mathematically aligns perfectly with the cooperative and symbiotic nature of the sports-tourism-agriculture nexus as observed in China.

Formally, the benevolent cross-efficiency evaluation proceeds in two stages. In the first stage, each DMU solves a standard CCR multiplier model to obtain its self-evaluated efficiency score and the corresponding optimal weight set. In the second stage, for each peer, re-optimizes its weights to maximize the efficiency of , subject to the constraint that its own efficiency remains at the self-evaluated optimum . The cross-efficiency of as evaluated by is mathematically defined as shown in Equation 1:

The final comprehensive cross-efficiency score for is then computed as the arithmetic mean of all peer evaluations as shown in Equation 2:where and denote the output and input weights optimized by under the benevolent objective; and are the -th output and -th input of ; and is the total number of .

This study uses the improvement of sports-tourism-agriculture coordinated development efficiency in Chinese provinces as the output indicator. Nine key indicators are identified as secondary metrics for industrial input across the three sectors within China, including the number of enterprises, workforce, and average annual wage (Zhao and Zhang, 2025; Zhang et al., 2025; Zhai et al., 2026). The detailed indicator system is presented in Table 1 for the Chinese provincial-level dataset.

2.1.2 Conditional variables

Table 2 summarizes the specific conditional variables used in this study for the Chinese provincial context. The operationalization and data sources for each condition are detailed below.

2.1.2.1 Digital technology innovation

As digital innovation has emerged as a core driving force of China’s economic development, we follow Ran et al. (2024) in using the volume of authorized patents from 2019 to 2023 to measure the provincial level of digital technology innovation in China.

2.1.2.2 Regional economic foundation

We operationalize this organizational-dimension condition using the provincial per capita GDP averaged over 2019–2023, drawn from the China Statistical Yearbook. Per capita GDP is widely adopted as a proxy for the regional economic foundation upon which cross-sectoral organizational coordination rests (Du et al., 2022). Within the TOE framework used in configurational studies of industrial integration (Su and Guo, 2023; Zhao and Zhang, 2025), regional economic foundation captures three organization-relevant dimensions: the material resources available for supporting cross-sectoral projects, the density of organizational and commercial infrastructure underpinning sports, tourism, and agricultural enterprises, and the aggregate capacity of regional actors to sustain integrated operations. This comprehensive quality makes per capita GDP particularly suitable as an organizational-dimension proxy. We acknowledge that per capita GDP does not directly measure government fiscal commitment or targeted policy intensity; however, in a balanced-panel design spanning 20 Chinese provinces over 5 years, it offers the methodological advantages of full data availability, cross-provincial comparability, and temporal consistency, all of which are essential for the NCA and fsQCA procedures employed here.

2.1.2.3 Cultural-tourism human resources

Local talent supply is proxied by the number of graduates from educational institutions affiliated with provincial cultural and tourism departments in 2019, sourced from the China Cultural Relics and Tourism Statistical Yearbook. We acknowledge that this indicator primarily captures cultural-tourism-related human capital rather than the full spectrum of human resources across all three sectors in China. This choice is driven by data availability—disaggregated human-capital statistics for the sports and agricultural sub-sectors are not systematically compiled at the provincial level in China during the sample period. The cultural-tourism workforce is nevertheless a reasonable proxy because it constitutes the principal integrative node connecting sporting events and agricultural activities through tourism within the Chinese rural tourism system.

2.1.2.4 Market demand

This condition is measured by the regional annual per capita consumption expenditure in Chinese provinces (Zhao and Zhang, 2025).

2.1.2.5 Resource endowment

A region’s resource base is assessed comprehensively through three dimensions: sports venue resources (Liu and Li, 2024; Zhao et al., 2015), the number of A-level scenic spots in China (Zhou et al., 2021), and the quantity and brand presence of specialty agricultural products across Chinese regions (Tan et al., 2024).

2.2 TOE research framework and system construction

The TOE framework has been widely used to examine how technological, organizational, and environmental conditions jointly shape digital transformation, innovation adoption, and cross-sector development. Recent studies have also combined the TOE framework with configurational methods such as fsQCA to analyze digital transformation pathways in agricultural and industrial contexts (Zheng et al., 2025; N’Dri and Su, 2024). In recent years, scholars have widely adopted the TOE framework theory to analyze industrial convergence practices and efficiency improvement scenarios in China, owing to its systematic decomposition of industrial development influencing factors and its multidimensional coverage. The theory is structured around three core dimensions: Technology (Wang and Yuan, 2025), Organization (Zhao and Zhang, 2025), and Environment (Su and Guo, 2023). Built on these three core dimensions—technological empowerment, organizational safeguards, and environmental factors—the analytical framework not only identifies technological enablers organizational safeguards, and external environmental support in industrial development, but also reveals the interactive dynamics among these elements through multidimensional variable analysis. Consequently, the framework provides a scientific and practical tool for understanding the efficiency mechanisms of complex cross-sectoral, multi-stakeholder participation models—such as the integration of sports, tourism, and agriculture in China—and for identifying key pathways to enhance coordinated development efficiency within China’s rural contexts.

This study adopts the TOE (Technology-Organization-Environment) framework as its meta-framework, integrating industrial synergy theory with China’s rural revitalization strategic objectives to establish an integrated theoretical foundation. The framework posits that enhancing the coordinated development efficiency of sports, tourism, and agriculture is a systemic project that serves the overarching narrative of China’s rural revitalization. Coordinated development efficiency is not determined by a single factor but hinges on the effective synergy among three key dimensions: technology, organization, and environment (Figure 1). Specifically, we select the level of digital technology innovation is selected. Digital technology innovation serves as the core driving force for high-quality development of China’s digital economy (Wan et al., 2024). The advancement of digital technology not only significantly facilitates the integration of sports (Li, 2024) and rural industries in China (Yan and Cao, 2024), but has also become a strategic choice for reshaping the tourism industry ecosystem and promoting its high-quality development (Chen et al., 2024). In the organizational dimension, regional economic foundation and cultural-tourism human resources are selected. The driving role of government policies and fiscal measures has been validated in multiple fields such as tourism (re)development (Zhang, 2011), sport sindustry (Hanwoong, 2019), and agriculture within the Chinese policy context (Desalegn et al., 2024). Additionally, cultural-tourism human resources have been confirmed as the key factor in industrial integration in China (Xue, 2025). In the environmental dimension, market demand and resource endowment conditions are selected. Market demand is the fundamental driving force for industrial development. It has been demonstrated through constructing a dynamic mechanism model and empirical analysis that market demand is the core driving factor for high-quality industrial integration in Chinese provinces (Li, 2025). To develop rural advantageous and distinctive industries and pursue a sustainable development path, it is also necessary to fully leverage the unique geographical conditions and resource endowments of rural areas, as argued by Chinese scholars (Yang et al., 2020), continuously developing new industries and business models within China’s rural territories. Furthermore, the “multiple concurrent” logic of this framework aligns closely with the fsQCA (fuzzy set qualitative comparative analysis) methodology, enabling us to transcend the traditional “net effect” mindset and to identify equivalent factors leading to high coordinated development efficiency in China. The condition configuration provides more targeted and practical strategic inspiration for practitioners in different regions and with different basic conditions across China.

Figure 1

3 Research design and data processing

3.1 Research methods

In identifying the driving mechanisms of industrial integration efficiency in China, previous studies have predominantly relied on traditional econometric models, such as multiple linear regression (OLS) or panel fixed-effects models. While these approaches are robust for estimating isolated variable impacts, they suffer from three major limitations when applied to complex systemic phenomena within Chinese rural contexts. First, traditional regressions assume symmetric causality, incorrectly positing that if the presence of a factor leads to high efficiency, its absence must inevitably lead to low efficiency. Second, they rely on a net-effects paradigm, assuming variables operate independently and linearly, thereby masking the complex synergistic interactions among multiple antecedent conditions. Third, they focus exclusively on average effects across a sample, failing to account for the reality of “equifinality”—the phenomenon where different regions of China, endowed with varying resources, can achieve the same high-efficiency outcomes through entirely different structural pathways.

Given that the integration of sports, tourism, and agriculture is a complex, multi-stakeholder ecosystem deeply embedded in the Technology-Organization-Environment (TOE) framework in China, this study replaces traditional linear regression with a configurational mixed-method approach combining NCA and fsQCA. This dual methodology mathematically addresses the limitations of standard models by distinguishing between “necessary bottlenecks” and “sufficient configurations,” capturing the asymmetric and equifinal nature of regional industrial integration across Chinese provinces.

3.1.1 Necessary condition analysis (NCA)

While standard regression estimates the average effect of a variable, NCA is designed to precisely identify “necessary but insufficient” conditions—the absolute bottleneck constraints in a dataset (Dul, 2016). NCA delineates a “ceiling line” in an X-Y scatter plot, quantifying the minimum threshold of condition required to achieve a specific level of outcome . This approach answers the critical question of whether certain resources or technological capabilities act as hard prerequisites for efficiency improvement in China, regardless of how favorably other factors are configured.

Formally, NCA quantifies the necessity of a condition for an outcome through the effect size , defined mathematically in Equation 3 as the ratio of the ceiling zone area to the total scope of the observation space:where denotes the area of the ceiling zone—the empty region above the ceiling line where no observations can exist if the condition is truly necessary—and represents the total empirical scope of the observation space. Statistical significance is evaluated through Monte Carlo permutation testing. An effect size d ≥ 0.1 with p < 0.05 is established as the standard threshold for a meaningful necessary bottleneck constraint.

3.1.2 Fuzzy set qualitative comparative analysis (fsQCA)

To complement NCA, fsQCA treats each provincial domain in China as a “configuration”—a holistic formula woven from multiple causative conditions (Ragin, 2008). Rooted in Boolean algebra and set theory, fsQCA overcomes the symmetric and net-effect limitations of traditional regression. It assumes complex causality, mathematically identifying multiple equivalent configurations of conditions (i.e., different pathways) that are sufficient to produce the same outcome within China’s provincial economies.

The empirical robustness of these causal configurations is evaluated using two core mathematical parameters: Consistency and Coverage.

Consistency measures the degree to which a specific configuration acts as a sufficient condition for the outcome (analogous to statistical significance). It is calculated as shown in Equation 4:

Coverage assesses the empirical relevance or explanatory power of the configuration, indicating the proportion of high-efficiency cases that exhibit this specific combination of conditions (analogous to in traditional regression). It is expressed mathematically as shown in Equation 5:where represents the calibrated fuzzy-set membership score of case in configuration , and represents its membership score in the outcome . By setting strict consistency thresholds (typically >0.75), fsQCA systematically reveals the context-dependent, multi-pathway drivers leading to high coordinated development efficiency in Chinese agriculture-tourism-sports systems.

3.2 Data sources and sample screening

This study aims to investigate the multiple concurrent factors influencing the coordinated development efficiency of sports, tourism, and agriculture in China, employing a hybrid method combining Necessary Condition Analysis (NCA) and Fuzzy Set Qualitative Comparative Analysis (fsQCA). To ensure the reliability of the analytical results, the completeness of the case samples and the quality of the data are crucial.

3.2.1 Raw data and preliminary processing

The raw data were collected from authoritative publications, including the China Statistical Yearbook, China Cultural Relics and Tourism Statistical Yearbook, China Sports Statistical Yearbook, and China Agricultural Statistical Yearbook, as well as the statistical yearbooks of various provinces in China and provincial government work reports. The collection period spans from 2019 to 2024, covering 30 provinces, autonomous regions, and municipalities directly under the central government in China (excluding Hong Kong, Macao, and Taiwan). We constructed a balanced panel dataset comprising 5 years and 150 “province-year” observations.

During panel construction, a small number of province–year cells were missing for certain indicators. To restore sample integrity while preserving the smoothness of year-on-year variation, we applied two complementary procedures as follows. Linear interpolation was used to close interior gaps in continuous series with stable trends, drawing on known values from adjacent years. Logarithmic trend fitting was used where systematic gaps or short-run volatility made adjacent-year interpolation unreliable, with missing values recovered from the fitted curve. Both procedures are standard for balanced panels of this type. All imputed cells are color-coded in the original dataset (red for linear interpolation, other colors for logarithmic fitting) to preserve full traceability.

To document the extent of imputation transparently, we report its exact share for every variable entering the final analytical sample of 1,900 province–year data points (20 Chinese provinces × 5 years × 19 variables). In total, 75 values (3.95%) were imputed. On the condition side, only 10 of 700 data points (1.43%) required treatment, all via linear interpolation, and these are confined to two of the seven indicators: sports venues (8 of 100) and the number of agricultural brands (2 of 100). The remaining five condition indicators—patents granted, Per capita GDP, number of graduates, per capita consumption expenditure, and number of A-level scenic spots—are fully observed. On the outcome side, 65 of 1,200 data points (5.42%) were imputed via logarithmic trend fitting, and all imputed values are located in the terminal year of the outcome panel, where official statistics had not yet been released at the time of data collection. The affected outcome indicators are sports employment (19 of 100), tourism employment (20 of 100), gross output of the sports industry (7 of 100), and gross output of the tourism industry (19 of 100); the remaining eight outcome indicators are fully observed. Across the 19 variables, nine contain no imputed values whatsoever, and the highest per-variable imputation share is 20%.

3.2.2 Sample screening process

Because both Necessary Condition Analysis (NCA) and fsQCA require every case (i.e., province-year observation) to possess complete and valid calibrated membership scores across all variables, we conducted rigorous data screening and cleaning. Eleven provinces were excluded due to missing data during the cleaning, calibration, or temporal lagging processes (specifically: Beijing, Jilin, Heilongjiang, Hainan, Gansu, Qinghai, Ningxia, Shanghai, Guizhou, Yunnan, and Xizang).

Our final sample comprises a balanced panel of 20 Chinese provinces (Anhui, Chongqing, Fujian, Guangdong, Guangxi, Hebei, Henan, Hubei, Hunan, Jiangsu, Jiangxi, Liaoning, Inner Mongolia, Shaanxi, Shandong, Shanxi, Sichuan, Tianjin, Xinjiang, and Zhejiang), yielding a total of 100 province-year observations. This sample demonstrates robust representativeness in terms of both geographic distribution and economic development levels across China, covering eight eastern, six central, and six western provinces (see Figure 2).

Figure 2

To account for the time-lag effect inherent in improving industrial coordinated development efficiency in Chinese provinces, our data structure pairs causal conditions from 2019 to 2023 with outcome variables from 2020 to 2024. Following established methodological precedents (e.g., Zhang et al., 2020), this design effectively introduces a one-year lag. The choice of a one-year interval is theoretically grounded in three substantive mechanisms specific to the sports-tourism-agriculture nexus in China.

First, China’s policy implementation predominantly operates on an annual budget and planning rhythm. Provincial policy documents issued in year t typically trigger resource allocation, project approval, and organizational deployment within the same fiscal year, with measurable output effects materializing in year t + 1 when newly funded programs complete their first full operational cycle. Second, the agricultural production system in China is inherently governed by annual seasonality. Planting, harvesting, and agri-tourism programming strictly follow calendar-year cycles, meaning that changes in input conditions during one growing season are most likely reflected in the subsequent year’s integrated output (Zhang et al., 2020). Third, digital technology adoption in the sports and tourism sectors in China follows a pattern of rapid deployment but delayed systemic impact. Platform launches, data accumulation, and user-base establishment typically require 6 to 12 months before generating stable efficiency gains in cross-sectoral service delivery (Chen et al., 2024; Gretzel et al., 2015). Taken together, these sector-specific temporal mechanisms converge on a one-year horizon as the most plausible interval for condition-to-outcome transmission.

3.3 Data calibration

To meet the requirements of fuzzy set qualitative comparative analysis (fsQCA), this study converts all raw data of conditional and outcome variables from Chinese provinces into membership scores within the [0, 1] interval. The conversion follows the “direct calibration method” to preserve ordinal information and prevent data loss, while establishing three theoretical qualitative anchor points for each variable: full membership (membership score 1), crossover point (membership score 0.5), and complete non-membership (membership score 0).

3.4 Principles for setting calibration anchor points

The calibration anchor points are established based on the principle of integrating data distribution characteristics with theoretical significance for the Chinese provincial sample. Complete membership anchor points are typically set at the 95th percentile, indicating cases that exhibit relatively high levels of the variable and can be considered “fully belonging” to the set. Cross-point anchor points are set at the median (50th percentile), representing cases in an intermediate state that neither fully belong nor fully belong to the set. Complete non-membership anchor points are set at the lower 5th percentile, indicating cases with relatively low levels of the variable and can be considered “fully non-belonging” to the set. The outcome variable is analyzed using the direct method. Fuzzy set calibration was performed, with 5, 50, and 95% quantiles serving as thresholds for complete non-membership, crossover points, and complete membership within China’s provincial data. The calibrated membership distribution was balanced, with no clustering of extreme values, thereby meeting the prerequisites for fsQCA analysis, as shown in Table 3.

4 Empirical analysis and discussion

4.1 Cross-efficiency analysis of benevolence

Using the R 4.5.2 benevolent cross-efficiency model to analyze the input–output data of sports, tourism, and agriculture in 20 provincial regions in China from 2020 to 2024, we obtained the annual coordinated development efficiency values of China’s provincial regions. From a temporal perspective, the average coordinated development efficiency of sports, tourism, and agriculture in China showed a steady upward trend during the sample period. Spatially, economically developed eastern coastal provinces in China such as Jiangsu, Zhejiang and Guangdong, as well as resource-rich central and western provinces of China like Sichuan and Hunan, maintained consistently high coordinated development efficiency levels. In contrast, some provinces in the northwest and northeast regions of China exhibited relatively lower efficiency values, reflecting significant regional disparities in coordinated development efficiency across China. Generally speaking, the regions in China with good economic foundation, abundant resources or flexible industrial coordination mechanisms have more outstanding performance in the coordinated development efficiency of sports, tourism and agriculture.

4.2 Necessary condition analysis

Table 4 presents the NCA results, revealing distinct patterns in the ceiling effects of individual conditions on coordinated development efficiency in Chinese provinces. Under the CE-FDH method, digital technology innovation (d = 0.305, p = 0.005) and resource endowment (d = 0.329, p = 0.040) demonstrate statistically significant bottleneck effects within China, indicating that these two conditions impose meaningful necessary constraints on achieving high coordinated development efficiency. Market demand shows a near-significant effect (d = 0.281, p = 0.060), suggesting a potential but not statistically confirmed bottleneck in the Chinese context. Regional economic foundation (d = 0.350, p = 1.000) and cultural-tourism human resources (d = 0.478, p = 0.814), despite exhibiting relatively large effect sizes, do not pass the significance test under the Monte Carlo permutation procedure, likely because of the distributional characteristics of the underlying data from Chinese provinces. Under the CR method, all conditions yield p-values close to or equal to 1.0, which is consistent with the known lower statistical power of the CR approach in NCA. Overall, although no factor serves as a singularly sufficient driver, the significant bottleneck effects (CE: p < 0.05) of digital innovation and resource endowment confirm their status as foundational necessary preconditions, imposing binding ceilings on the maximum achievable efficiency in China. Table 5 reports the bottleneck level analysis derived from the NCA results. It shows the minimum level that each antecedent condition needs to reach, where applicable, for coordinated development efficiency to attain different target levels among Chinese provinces. To achieve 100% coordinated development efficiency in Chinese provinces, the two statistically significant bottleneck conditions—digital technology innovation and resource endowment—must reach at least 81.8 and 77.2%, respectively. The thresholds for regional economic foundation and cultural-tourism human resources are not statistically significant and should not be interpreted as mandatory requirements. Market demand shows a marginally significant bottleneck effect (p = 0.060), requiring a minimum level of 89.0% to attain the maximum efficiency level. For a target of 40% coordinated development efficiency, only digital technology innovation and resource endowment act as essential preconditions, while other factors are non-essential. This finding implies that, in the initial stage of efficiency improvement across China, technological empowerment and resource endowment play more fundamental supporting roles.

The results in Table 5 should be interpreted in light of the statistical significance of the NCA effects reported in Table 4. Digital technology innovation and resource endowment demonstrate significant ceiling effects under the CE-FDH method (p < 0.05), indicating that their bottleneck thresholds have substantive interpretive value. By contrast, regional economic foundation and cultural-tourism human resources do not reach statistical significance, while market demand is only marginally significant (p < 0.1). Therefore, the thresholds for these latter conditions should be interpreted cautiously as indicative rather than confirmatory.

4.3 Single condition necessity test

The single-condition necessity analysis (see Supplementary Table S1) reveals that all condition variables demonstrate consistency indices below the 0.9 threshold. This indicates that, while certain conditions were identified as bottlenecks in the previous NCA, they do not constitute absolute necessary conditions in a set-theoretic sense for high coordinated development efficiency. These findings further confirm that achieving high efficiency exhibits typical causal complexity, characterized by “multiple conjunctural causation”. No single condition is sufficient on its own to independently explain high efficiency, necessitating a comprehensive analysis from a configurational perspective. Building on this, we further examine potential causal relationship combinations with adjusted distances exceeding 0.2. All combinations showed consistency below the 0.9 threshold, indicating no significant necessity relationship in Chinese provinces. This finding complements the NCA bottleneck analysis. While the NCA identifies specific conditions as necessary bottleneck constraints at certain efficiency levels, the fsQCA results clarify that none of these factors are universally required across all high-efficiency cases. This underscores that efficiency enhancement stems from the synergistic interaction of multiple conditions under varying contexts rather than from a single “must-have” factor.

4.4 Conditional sufficiency analysis

Building upon the validation that no single necessary condition exists, this study further employs fsQCA to explore the sufficiency pathways of multi-factor collaborative effects on the coordinated development efficiency of sports, tourism, and agriculture from a configurational perspective in China. During the analysis, the case frequency threshold was set at 2 and the consistency threshold at 0.87 (Schneider and Wagemann, 2012; Greckhamer et al., 2018). The variables were converted into fuzzy set membership degrees using the direct calibration method. Ultimately, we identified three condition configurations that contribute to high coordinated development efficiency, with an overall consistency of 0.813. The overall coverage was 0.570, indicating that these configurations demonstrated strong explanatory power for high-efficiency cases across Chinese provinces (Li and Luo, 2026; Wu et al., 2025).

4.4.1 Configuration path types and mechanisms

This study yields two types of fsQCA solutions: the parsimonious solution (see Supplementary Table S2) and the intermediate solution. The intermediate solution, which incorporates theoretically justified directional expectations and serves as the primary analytical basis, is presented in Table 6. Following established fsQCA best practices (Ragin, 2008; Fiss, 2011), core conditions are identified as those appearing in both the parsimonious and intermediate solutions, while peripheral conditions are those present only in the intermediate solution. Table 7 synthesizes these results by displaying the three distinct configurational pathways derived from the intermediate solution, with core and peripheral conditions clearly distinguished. Employing Boolean minimization algorithms in fsQCA 3.0 software, this study identified three distinct configuration paths leading to high coordinated development efficiency across sports, tourism, and agriculture (Table 7). The overall solution consistency reaches 0.813, surpassing the commonly accepted threshold of 0.75, thereby confirming the robustness of the configurational causal relationships. The solution coverage of 0.570 indicates that these four configurations collectively account for 57.0% of the membership in high-efficiency cases, reflecting strong empirical representativeness.

4.4.1.1 Configuration 1: resource endowment driven

Following established fsQCA best practices (Ragin, 2008), this study adopts the intermediate solution as the primary analytical basis. Core conditions are identified as those conditions appearing in both the parsimonious and intermediate solutions, while peripheral conditions are those present only in the intermediate solution. This distinction, consistent with the configurational approach advocated by Fiss (2011), enables a theoretically grounded interpretation of each pathway’s causal mechanism.

4.4.1.1.1 Configuration 1: resource endowment–driven path

This configuration is defined by the conjunction of three core conditions: the absence of regional economic foundation (~regional economic foundation), the absence of market demand (~market demand), and the presence of resource endowment. It yields the highest unique coverage among the three paths (0.101), with a raw coverage of 0.481 and a consistency of 0.872, indicating strong explanatory distinctiveness. Notably, all constituent conditions are core conditions, with no peripheral factors supplementing this pathway.

This configuration reveals that in contexts characterized by limited governmental intervention and subdued market demand, abundant and distinctive natural and cultural resource endowments can independently serve as the primary driver of cross-sector integration. This finding resonates with the resource-based view of competitive advantage (Barney, 1991), which posits that rare, valuable, and inimitable resources constitute the foundation of sustained superior performance. Regions conforming to this path typically possess high-quality sports facilities, nationally recognized scenic areas, or distinctive agricultural brands whose inherent appeal and competitive advantage compensate for institutional and market deficiencies. The causal logic suggests that when external policy incentives and local consumer purchasing power are weak, the intrinsic attractiveness of exceptional resources can still attract external visitors and investment, thereby generating coordinated development outcomes independently.

Empirically, this pathway finds resonance in resource-dependent rural development contexts documented in the literature. Fleischer and Tchetchik (2005) demonstrated that rural tourism operations in Israel significantly benefit from proximity to unique natural and agricultural attractions, even absent coordinated public policy. In the Chinese context, Xinjiang has capitalized on its vast grasslands, snow-capped mountains, and ethnic sports culture to develop initiatives such as the Saiyimu Lake Cycling Race and Kanas Hiking Tourism programs. Despite relatively limited policy support and modest local consumer markets, the region’s exceptional resource base has enabled high coordinated development efficiency. Similarly, Fujian has successfully integrated tea culture with ecotourism through projects such as the Wuyishan Tea-Tourism Integration Demonstration Zone (Shangguan et al., 2022), exemplifying the value of a resource-endowment-driven integration approach.

4.4.1.1.2 Configuration 2: technology–market synergy path

This configuration features digital technology innovation and the absence of regional economic foundation (~regional economic foundation) as core conditions, supplemented by market demand as a peripheral condition. It achieves a raw coverage of 0.336, a unique coverage of 0.021, and a consistency of 0.844. The causal logic embedded in this path suggests that the combination of technological innovation capacity and active market mechanisms can effectively substitute for formal policy support in driving coordinated development efficiency. This substitution effect is theoretically consistent with the equifinality principle central to configurational analysis (Ragin, 2008), where in distinct combinations of conditions can produce functionally equivalent outcomes.

The pathway aligns with growing evidence that digital technologies serve as transformative enablers of industrial convergence. Global evidence indicates that digital technology adoption drives the integration of sports and health industries (Li, 2024). Furthermore, digital innovation reshapes the tourism industry ecosystem by enabling demand-responsive service design and platform-mediated resource aggregation (Chen et al., 2024). The peripheral role of market demand in this configuration suggests that, while robust consumer markets enhance the pathway’s efficacy, digital technology and institutional decentralization constitute its essential causal core.

Regions exhibiting this configuration typically possess advanced digital infrastructure alongside robust consumer markets, enabling market-responsive technological applications to spontaneously generate integrated business models. For example, Jiangsu Province has developed the Sports Tourism Consumption Season initiative, wherein events such as the Rural Leisure Sports Festival leverage both digital platforms and the province’s strong consumer base to foster deep convergence among sports competitions, rural tourism, and agricultural product exhibitions. Guangdong Province similarly exemplifies this pathway by capitalizing on the high consumption capacity of the Pearl River Delta region to develop digitally enhanced integrated projects such as the Shenzhen Marathon combined with rural tourism experiences and the Guangzhou Fitness Leisure and Eco-Farm initiatives, demonstrating how the synergy between digital technology and market demand powerfully enhances coordinated development efficiency—even without direct government intervention.

4.4.1.1.3 Configuration 3: digital innovation–resource compensation path

This configuration is characterized by the conjunction of digital technology innovation, the absence of cultural-tourism human resources (~cultural-tourism human resources), and resource endowment—all identified as core conditions. It achieves a raw coverage of 0.414, a unique coverage of 0.058, and a consistency of 0.866. This pathway reveals that digital technology can serve as a critical instrument for “leapfrog” development: under conditions of relative human capital scarcity, the combination of technological empowerment and natural resource advantages can substantially enhance coordinated development efficiency.

This finding extends the emerging literature on digital compensation mechanisms in rural development. Rural industrial integration is facilitated by digital economy development, which reduces transaction costs and information asymmetries and effectively compensates for institutional and human capital constraints (Yan and Cao, 2024). Similarly, the capacity of digital technology innovation to bridge regional disparities in factor endowments is documented through its spatial and temporal dynamics in China (Wan et al., 2024). In this configuration, digital platforms perform functions—such as intelligent recommendation, virtual experience design, and automated service delivery—that would otherwise require specialized human capital, thereby enabling resource-rich but talent-scarce regions to achieve high coordinated-development outcomes.

Regions conforming to this path typically exhibit comparative advantages in digital infrastructure and innovation platforms, enabling resource virtualization, intelligent service delivery, and platform-based operations that effectively compensate for talent shortages. For example, Zhejiang Province has leveraged its Digital Village and Smart Tourism platforms to launch initiatives such as Cloud Marathon events and VR Agricultural Experience projects, breaking down the physical barriers between sports, tourism, and agricultural sectors through digital means. Similarly, Shandong Province has leveraged its advanced digital infrastructure and rich agricultural resource base to deploy big data analytics and e-commerce platforms that connect sports event organization, rural tourism route curation, and specialty agricultural product distribution—demonstrating that digital technology innovation, when combined with available resource endowments, can overcome the constraints imposed by limited specialized human capital.

4.4.2 Regional compatibility analysis

Building on the three identified configuration paths, this study further examines their spatial distribution characteristics across China’s eastern, central, and western regions. The analysis reveals that different pathways have formed clear matching relationships with the resource conditions, market foundations, and technological environments of each macro-region, presenting a systematic pattern of “adapting to local conditions and path differentiation” (as shown in Figure 3).

Figure 3

The eastern region primarily aligns with Configuration 2 (Technology–Market Synergy) and Configuration 3 (Digital Innovation–Resource Compensation). This area possesses two core strengths: first, robust market consumption capacity with leading per capita disposable income and expenditure levels, creating a stable and multi-tiered demand base for integrated sports-tourism-agriculture products; second, a mature digital economy supported by well-developed information infrastructure and vibrant technological innovation. These advantages enable the region to leverage market mechanisms for resource allocation and to drive industrial upgrading through digital empowerment, forming a dual-driven model of “market traction—technology empowerment.” Provinces such as Jiangsu, Zhejiang, and Guangdong exemplify this pattern, having established smart cultural tourism platforms and digitally enhanced consumption scenarios that achieve efficient alignment between demand and technology (Guo et al., 2020).

The central region demonstrates a more complex alignment pattern. Certain resource-rich provinces exhibit characteristics consistent with Configuration 1 (Resource Endowment–Driven), while others with concentrated educational and scientific research resources show potential alignment with Configuration 3 where digital infrastructure investment is expanding. This region possesses abundant natural and cultural resources, though the efficiency of converting these resources into value remains suboptimal. Enhancing coordinated development efficiency hinges on leveraging either exceptional resource endowments or emerging digital capabilities to compensate for relatively constrained market scale. Hunan Province, for instance, has capitalized on its distinctive natural landscapes and cultural heritage to develop integrated sports-tourism products, while simultaneously investing in digital platforms to improve resource management and visitor experience (Zhao and Zhang, 2025).

The western region is the core area of Configuration 1 (Resource Endowment–Driven Path). Despite constraints in local market size and infrastructure, this region boasts unique and irreplaceable natural and cultural resources—snow-covered plateaus, ethnic customs, and vast wilderness landscapes—which constitute its core comparative advantages. The NCA results confirm that resource endowment exhibits statistically significant bottleneck effects (CE: p = 0.040), underscoring its foundational constraining role. Regions such as Xinjiang and Sichuan leverage their top-tier resources to develop distinctive sports events and in-depth tourism routes, successfully attracting long-distance visitors and realizing a development path of “trading resources for the market and shaping brand with characteristics.”

To rigorously validate the regional differentiation patterns identified above, we conducted a cross-tabulation analysis using Fisher–Freeman–Halton Exact Test (Freeman and Halton, 1951; Fagerland et al., 2017) was conducted. Each of the 19 analytically valid provinces was assigned to its dominant configuration pathway based on the higher intermediate solution coverage value. Shanxi province was excluded from this analysis as it does not attain high coordinated efficiency and therefore lacks a valid dominant pathway assignment. For provinces exhibiting dual configurational alignment as indicated in Figure 3, the pathway with the higher intermediate solution coverage was designated as the primary assignment, yielding mutually exclusive and exhaustive province-to-pathway assignments across the sample. The frequency distribution reveals a structurally differentiated pattern across the three macro-regions (see Supplementary Table S3). Eastern provinces are exclusively concentrated in the Technology–Market Synergy path (C2: 62.5%) and the Digital Innovation–Resource Compensation path (C3: 37.5%), with no province assigned to the Resource Endowment–Driven path (C1: 0.0%). Central provinces exhibit a predominantly resource-endowment-oriented structure (C1: 80.0%), with limited representation in C3 (20.0%) and no presence in C2. Western provinces display complete concentration in the Resource Endowment–Driven path (C1: 100.0%), with no representation in either C2 or C3. Given that all expected cell counts fall below 5 (minimum expected count = 1.05), the Fisher–Freeman–Halton Exact Test was applied in preference to the asymptotic Pearson Chi-square, as the exact procedure does not rely on large-sample assumptions. The results are statistically decisive: Fisher’s exact p < 0.001, with a Cramér’s V = 0.662, indicating a large-magnitude association between geographic macro-region and dominant integration pathway. These results formally reject the null hypothesis of random distribution and elevate the regional compatibility argument from empirical observation to robust statistical inference, confirming that regional factor endowments fundamentally dictate the optimal integration pathways in the sports-tourism-agriculture nexus.

Importantly, the dominant paths in different regions are not isolated from each other. Against the backdrop of increasingly smooth factor flows, cross-regional path mutual learning and collaboration have become feasible (Fan et al., 2012). Digital technologies and management models from the eastern regions can empower resource development in the central and western regions; distinctive products from the west can expand sales through the market networks of the east. This cross-regional complementarity, consistent with previously documented spatial spillover effects (Zhao et al., 2015), helps transcend the limitations of single-regional paths and optimize resource allocation on a larger scale. Regional development strategies should therefore be tailored to local conditions, with each area selecting optimal pathways while establishing regional coordination mechanisms to foster a synergistic ecosystem for integrated development.

4.4.3 Robustness test

To address the possibility that our configurational findings depend on the choice of consistency threshold, we re-ran the fsQCA analysis at two alternative specifications while holding all other parameters constant (frequency cutoff = 2; calibration anchors at 5/50/95 percentiles). Specifically, we tested a more lenient threshold of 0.85 and a more stringent threshold of 0.90, thereby bracketing our main-analysis value of 0.87 from both directions (see Supplementary Table S4). Three findings support the robustness of our results. First, all three theoretically substantive configurations identified in the main analysis retain identical core-condition structures across the 0.85 and 0.87 specifications. At the 0.85 threshold, the Resource-Endowment–Driven path, the Technology–Market Synergy path, and the Digital-Innovation–Resource-Compensation path appear with exactly the same conjunctions as in the main analysis. An additional fourth path emerges at 0.85, but its unique coverage (0.016) falls below the 0.02 empirical triviality benchmark identified by Schneider and Wagemann (2012, p. 135) and was therefore not retained in the principal solution. Second, the direction of every core condition remains invariant across all three specifications—no condition reverses its polarity when the threshold is raised or lowered, indicating that the substantive causal interpretation of each pathway is stable. Third, although Configuration 3 is not retained at the 0.90 threshold, this change is a predictable consequence of adopting a threshold at the upper edge of the range recommended by Greckhamer et al. (2018) and of accepting a solution coverage of 0.466, which falls below the 0.50 benchmark generally regarded as indicating meaningful empirical scope (Schneider and Wagemann, 2012, p. 129). The elimination of Configuration 3 therefore reflects a reduction in empirical coverage rather than a challenge to its substantive validity; its core-condition structure, where it is retained, is identical to the main analysis.

Taken together, the core-condition composition of all three theoretically substantive configurations remains stable across a 0.05-unit span of consistency thresholds (0.85 to 0.90), providing evidence that the identified configurational mechanisms are not artifacts of a particular parameter choice.

5 Conclusions and recommendations, limitations and future directions

5.1 Conclusion

Based on the TOE theoretical framework, this study uses balanced panel data from 20 Chinese provinces (2020–2024) and combines NCA with fsQCA to analyze the multiple concurrent paths for improving the coordinated development efficiency of sports-tourism-agriculture integration in China. Three principal findings emerge.

First, digital technology innovation and resource endowment impose statistically significant “necessary-but-insufficient” bottleneck constraints, yet no single condition is individually sufficient. NCA identifies digital technology innovation and resource endowment as the only two conditions exhibiting significant ceiling effects, while the fsQCA necessity test confirms that no condition reaches the set-theoretic necessity threshold. This convergence between bottleneck and configurational evidence substantiates the “multiple conjunctural causation” logic of configurational analysis.

Second, three equivalent configurational pathways drive high coordinated efficiency through substitution and complementarity mechanisms: (a) the Resource Endowment–Driven Path, where exceptional natural and cultural resources compensate for weak policy and market conditions; (b) the Technology–Market Synergy Path, where digital innovation combined with consumer demand substitutes for formal regional economic foundation; and (c) the Digital Innovation–Resource Compensation Path, where technological empowerment compensates for human-capital scarcity. Resource endowment and digital technology innovation recur as core conditions in two of three configurations, demonstrating that the strategic value of resource endowments is reconfigured—rather than diminished—by digital technologies in cross-sectoral integration contexts within China.

Third, the three pathways exhibit distinct regional adaptability that mirrors China’s spatial development patterns. Eastern coastal provinces predominantly align with the technology–market synergy and digital innovation–resource compensation paths; western regions follow the resource endowment–driven path; central regions exhibit mixed alignment reflecting their transitional factor endowments. Cross-regional complementarity between eastern digital capabilities and western distinctive resources constitutes a significant avenue for collaborative development.

5.2 Recommendations

Based on the three identified pathways, this study proposes differentiated policy recommendations from three dimensions: technology, organization, and environment, aiming to build a tripartite integration promotion system for different regions and paths in China.

5.2.1 Technical dimension: driving convergent innovation through digitalization and intelligence

For the Resource Endowment–Driven path, lightweight digital tools—smart navigation, VR-based cultural experiences, and blockchain-enabled agricultural traceability—should serve as value amplifiers that preserve ecological and cultural integrity. For the Technology–Market Synergy path, big data analytics and AI-driven personalization should anchor consumption upgrades, supported by robust data governance frameworks. For the Digital Innovation–Resource Compensation path, intelligent substitution technologies—AI-powered tour guides, automated agricultural management, and digital service delivery—should be prioritized to compensate for specialized human-capital scarcity, supplemented by integrated innovation hubs that accelerate patent commercialization.

5.2.2 Organizational dimension: stimulating endogenous motivation through institutional synergy and talent support

For the Resource Endowment–Driven path, organizational focus should rest on guardrail-type regulations—heritage protection, resource valuation, and community benefit-sharing mechanisms—rather than interventionist industrial policy. For the Technology–Market Synergy path, the optimal posture is enabling rather than directive: accelerated digital infrastructure construction, streamlined market entry for cross-sectoral enterprises, and innovation-oriented consumption funds. For the Digital Innovation–Resource Compensation path, priority should be given to collaborative government–university–enterprise education systems, specialized training bases for cross-sectoral integration professionals, and competitive talent-attraction packages.

5.2.3 Environmental dimension: policy guidance and market development to optimize external support

For the Resource Endowment–Driven path, external market development is critical: targeted marketing towards developed-region consumers, combined with partnerships with online travel agencies, sports media platforms, and agricultural e-commerce channels. For the Technology–Market Synergy path, environmental strategy should deepen online-offline synergy through digital consumption formats—cloud sports events, VR-enhanced agricultural experiences, smart navigation systems—while cultivating emerging segments such as family-oriented sports tourism. For the Digital Innovation–Resource Compensation path, systematic brand development and cross-regional market expansion should consolidate regional resource advantages through signature itineraries, brand alliances, and competitive event-bidding mechanisms.

5.3 Limitations and future directions

Three limitations warrant acknowledgment and provide avenues for future research.

First, the outcome variable captures coordinated sectoral efficiency rather than pure integration synergy. Therefore, future research should develop purpose-built integration indicators, such as inter-sectoral input–output linkages or shared-customer overlap metrics. Relatedly, future studies could combine document-based and expenditure-based indicators to more accurately capture targeted policy intensity.

Second, the exclusion of 11 provinces may limit the generalizability of the configurational findings. As disaggregated data become available, future studies should incorporate these provinces to test whether the identified pathways generalize across the complete national sample, ideally paired with multi-period lag designs or dynamic QCA approaches.

Third, the calibration thresholds are data-driven rather than theoretically derived. Theoretically grounded anchors—derived from official benchmarks, national policy targets, or expert consensus—would enhance the external validity of set-theoretic claims. Future research should explore mixed calibration strategies that combine percentile-based and substantively anchored thresholds.

References

• 1

  BankerR. D.CharnesA.CooperW. W. (1984). Some models for estimating technical and scale inefficiencies in data envelopment analysis. Manag. Sci.30, 1078–1092. doi: 10.1287/mnsc.30.9.1078,

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 2

  BarneyJ. (1991). Firm resources and sustained competitive advantage. J. Manage.17, 99–120. doi: 10.1177/014920639101700108

  • CrossRef
  • Google Scholar

• 3

  BennerM. J.RanganathanR. (2013). Divergent reactions to convergent strategies: investor beliefs and analyst reactions during technological change. Organ. Sci.24, 378–394. doi: 10.1287/orsc.1120.0755,

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 4

  CharnesA.CooperW. W.RhodesE. (1978). Measuring the efficiency of decision making units. Eur. J. Oper. Res.2, 429–444. doi: 10.1016/0377-2217(78)90138-8

  • CrossRef
  • Google Scholar

• 5

  ChenX.LiL.LingX. (2024). Research on driving factors of digital innovation in the tourism industry. SAGE Open14, 1–15. doi: 10.1177/21582440241304537

  • CrossRef
  • Google Scholar

• 6

  DesalegnG.TanglA.Fekete-FarkasM.GudisaG.BorosA. (2024). Linking policies and regulations to sustainable finance for the promotion of urban agriculture: evidence from micro and small businesses. Heliyon10:e31938. doi: 10.1016/j.heliyon.2024.e31938,

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 7

  DorockiS.SzymańskaA. I.Zdon-KorzeniowskaM. (2016). “Agricultural tourism farms in Poland: how the farmers improve their businesses—a case study,” in Understanding Innovation in Emerging Economic Spaces, ed. MicekG. (London: Routledge), 247–262.

  • Google Scholar

• 8

  DoyleJ.GreenR. (1994). Efficiency and cross-efficiency in DEA: derivations, meanings and uses. J. Oper. Res. Soc.45, 567–578. doi: 10.1057/jors.1994.84

  • CrossRef
  • Google Scholar

• 9

  DuY. Z.LiuQ. C.ChenK. W. (2022). Multiple models of business environment ecology, total factor productivity, and urban high-quality development: a configurational analysis based on complex systems view. Manage. World38, 127–145. doi: 10.19744/j.cnki.11-1235/f.2022.0123

  • CrossRef
  • Google Scholar

• 10

  DulJ. (2016). Necessary condition analysis (NCA): logic and methodology of “necessary but not sufficient” causality. Organ. Res. Methods19, 10–52. doi: 10.1177/1094428115584005

  • CrossRef
  • Google Scholar

• 11

  FagerlandM. W.LydersenS.LaakeP. (2017). Statistical Analysis of Contingency Tables. London: Chapman and Hall/CRC.

  • Google Scholar

• 12

  FanJ.SunW.ZhouK.ChenD. (2012). Major function oriented zone: new method of spatial regulation for reshaping regional development pattern in China. Chin. Geogr. Sci.22, 196–209. doi: 10.1007/s11769-012-0528-y

  • CrossRef
  • Google Scholar

• 13

  FarrellM. J. (1957). The measurement of productive efficiency. J. R. Stat. Soc. Ser. A Stat. Soc.120, 253–281. doi: 10.2307/2343100

  • CrossRef
  • Google Scholar

• 14

  FissP. C. (2011). A set-theoretic approach to organizational configurations. Acad. Manag. Rev.36, 135–160. doi: 10.5465/amr.2007.26586092

  • CrossRef
  • Google Scholar

• 15

  FleischerA.TchetchikA. (2005). Does rural tourism benefit from agriculture?Tour. Manag.26, 493–501. doi: 10.1016/j.tourman.2003.10.003

  • CrossRef
  • Google Scholar

• 16

  FreemanG. H.HaltonJ. H. (1951). Note on an exact treatment of contingency, goodness of fit and other problems of significance. Biometrika38, 141–149. doi: 10.1093/biomet/38.1-2.141,

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 17

  GreckhamerT.FurnariS.FissP. C.AguileraR. V. (2018). Studying configurations with qualitative comparative analysis: best practices in strategy and organization research. Strateg. Organ.16, 482–495. doi: 10.1177/1476127018786487

  • CrossRef
  • Google Scholar

• 18

  GretzelU.SigalaM.XiangZ.KooC. (2015). Smart tourism: foundations and developments. Electron. Mark.25, 179–188. doi: 10.1007/s12525-015-0196-8

  • CrossRef
  • Google Scholar

• 19

  GuoY.LiuH.ChaiY. (2020). The embedding convergence of digital economy and traditional industries: theoretical analysis and empirical evidence. Econ. Res. J.55, 1–16.

  • Google Scholar

• 20

  HanwoongY. (2019). Government involvement for the promotion of sport industry and the economic impact: application of input-output analysis to sport industry supporting business. Korean J. Phys. Educ.58, 387–404. doi: 10.23949/kjpe.2019.03.58.2.387

  • CrossRef
  • Google Scholar

• 21

  KimN.LeeH.KimW.LeeH.SuhJ. H. (2015). Dynamic patterns of industry convergence: evidence from a large amount of unstructured data. Res. Policy44, 1734–1748. doi: 10.1016/j.respol.2015.02.001

  • CrossRef
  • Google Scholar

• 22

  LiJ. (2024). Can digital technology drive the integration of sports and health industries? Evidence from global businesses. J. Public Health46, e697–e699. doi: 10.1093/pubmed/fdae169,

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 23

  LiZ. (2025). The driving forces and paths for the high-quality integration development of Guangdong’s culture, sports, and tourism under the “dual circulation” background. Sci. Technol. Station. Sport. Goods, 49–52.

  • Google Scholar

• 24

  LiS.LiangL.HuangZ.WangQ.SunY. (2025). Spatiotemporal patterns and driving forces of net primary productivity in south and southeast Asia based on Google earth engine and MODIS data. Catena249:108865. doi: 10.1016/j.catena.2024.108689

  • CrossRef
  • Google Scholar

• 25

  LiX.LuoX. (2026). The theoretical mechanism and multiple paths of digital ecosystems empowering urban carbon reduction: analysis of complex mediation effects based on dynamic QCA. J. Nat. Resour.41, 479–499. doi: 10.31497/zrzyxb.20260210

  • CrossRef
  • Google Scholar

• 26

  LiuJ.LiH. (2024). Research on the coupling relationship between sports resources and tourism industry from the perspective of sports and tourism integration—a case study of the three northeastern provinces. J. Liaoning Univ. (Philos. Soc. Sci. Ed.)52, 65–77.

  • Google Scholar

• 27

  N’DriA. B.SuZ. (2024). Successful configurations of technology–organization–environment factors in digital transformation: evidence from exporting small and medium-sized enterprises in the manufacturing industry. Inf. Manag.61:104030. doi: 10.1016/j.im.2024.104030

  • CrossRef
  • Google Scholar

• 28

  QiuH. (2025). Striving to enhance comprehensive agricultural production capacity and quality efficiency. Issues. Agricult. Econ., 18–22. doi: 10.13246/j.cnki.iae.2025.12.007

  • CrossRef
  • Google Scholar

• 29

  RaginC. C. (2008). Redesigning Social Inquiry: Fuzzy Sets and Beyond. Chicago, IL: University of Chicago Press.

  • Google Scholar

• 30

  RanL.ZhangQ.LiX. (2024). Can digital technology innovation contribute to firms’ market value?PLoS One19:e0309993. doi: 10.1371/journal.pone.0309993,

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 31

  SchneiderC. Q.WagemannC. (2012). Set-Theoretic Methods for the Social Sciences: A Guide to Qualitative Comparative Analysis. Cambridge: Cambridge University Press.

  • Google Scholar

• 32

  ShangguanC.HuangZ.HuangS. (2022). Thoughts on the digital transformation development of Wuyishan tea tourism town in the era of big data. China Manag. Information.25, 169–172.

  • Google Scholar

• 33

  SuZ.GuoL. (2023). How to improve the efficiency of cultural and tourism integration?—a study on configurational effects based on the TOE framework. J. Guangxi Univ. (Philos. Soc. Sci. Ed.)45, 122–130.

  • Google Scholar

• 34

  TanS.LiuJ.ZhaoY.XiC. (2024). Evaluation and empirical analysis on the development index of the characteristic agricultural in Yunnan county level. J. Yunnan Agricult. Univ. (Soc. Sci.)18, 75–78.

  • Google Scholar

• 35

  WanX.WangY.ZhangW. (2024). The spatial and temporal situation of China's digital technology innovation and its influencing factors. PLoS One19:e0297401. doi: 10.1371/journal.pone.0297401,

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 36

  WangR.YuanQ. (2025). Configurational research on integration of regional cultural industry and tourism industry—an exploration based on fsQCA. Tourism Science39, 38–58.

  • Google Scholar

• 37

  WangY.ZhangQ.ShaoY. (2022). Situation measurement and influencing factors of the coupling and coordination between sports industry and tourism industry in China. J. Shandong Sport Univ.38, 43–53.

  • Google Scholar

• 38

  WuY.HeZ.ZhouH. (2025). Multiple cultivation pathways for new quality productive forces in provincial-level regions: from the perspective of digital innovation ecosystem. J. Manag.38, 31–46.

  • Google Scholar

• 39

  WuJ.LiuP. (2022). Research on the integration and development of “sports & tourism agriculture” under the strategy of rural revitalization. Theory. Practice. Finance. Econ.43, 148–154.

  • Google Scholar

• 40

  XiangH.ZhouC.DongC.SongH. (2024). High-quality agricultural development in the Central China: empirical analysis based on the Dongting Lake area. Geomatica76:100010. doi: 10.1016/j.geomat.2024.100010

  • CrossRef
  • Google Scholar

• 41

  XuX.CaoB.XuR.LiuY. (2025). The logical mechanism, functional mechanism, and realization path of China’s outdoor sports industry empowering regional integration of production, living, and ecology. J. Phys. Educ.32, 69–78. doi: 10.16237/j.cnki.cn44-1404/g8.2025.06.006

  • CrossRef
  • Google Scholar

• 42

  XueW. (2025). Research on the linked development and integrated development model of sailing and windsurfing industry and tourism resources from the perspective of Hainan free trade port. In Proceedings of the 2025 Third Sichuan Sports Science Conference Paper Presentation. (Beijing: Beijing Sport University), 1, 69–73. doi: 10.26914/c.cnkihy.2025.037172

  • CrossRef
  • Google Scholar

• 43

  YanM.CaoX. (2024). Digital economy development, rural land certification, and rural industrial integration. Sustainability16:4640. doi: 10.3390/su16114640,

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 44

  YangY.LuoQ.LunR.ZhangQ.ChengC. (2020). Research on the development dynamic mechanism of rural advantageous characteristic industry from the perspective of system analysis. Issues. Agricult. Econ., 61–73.

  • Google Scholar

• 45

  ZhaiF.ShiT.NiuD. (2026). Research on the mechanism and influencing factors of sports-tourism integration in the Yellow River Basin under the context of digital economy. J. Henan Normal Univ. (Nat. Sci. Ed.)54, 54–64.

  • Google Scholar

• 46

  ZhangL. (2011). Research on fiscal and taxation policies promoting the development of China’s tourism industry [Master’s thesis, Hebei University]. Baoding: Hebei University.

  • Google Scholar

• 47

  ZhangG.LiuW.BaoH. (2020). Study on the time-lag effect of multiple environmental regulation policies on regional industrial structure change. J. Manag. Sci. China23, 95–107.

  • Google Scholar

• 48

  ZhangY.ZhengR.WangZ. (2025). The dynamic evolution of the integration development level of China’s sports, culture and tourism industries and its driving factors. J. Xi’an Phys. Educ. Univ.42, 308–326.

  • Google Scholar

• 49

  ZhaoJ.ZhangW. (2025). Pathways to improve the integration efficiency of regional sports and tourism industry: NCA and dynamic QCA analysis based on TOE framework. J. Sports Res.39, 58–71.

  • Google Scholar

• 50

  ZhaoZ.ZhangX.ShenN. (2015). The multi-dimensional spillover effects of regional collaborative innovation efficiency. China Ind. Econ., 32–44. doi: 10.19581/j.cnki.ciejournal.2015.01.003

  • CrossRef
  • Google Scholar

• 51

  ZhengY.LiaoF.TianM. (2025). Examining the factors influencing the digital transformation of new agricultural operating entities: insights from Zhejiang, China. Humanit. Soc. Sci. Commun.12:608. doi: 10.1057/s41599-025-04949-y

  • CrossRef
  • Google Scholar

• 52

  ZhouH.ZhangY.NingX.ZangS. (2021). Spatial-temporal differentiation characteristics of A-class tourist attractions in Inner Mongolia. J. Arid Land Resour. Environ.35, 202–208.

  • Google Scholar