Spatiotemporal Evolution and Spillover Effects of Tourism Industry and Inclusive Green Growth Coordination in the Yellow River Basin: Toward Sustainable Development

Abstract

Balancing tourism industry (TI) growth and ecological protection is critical for sustainable development in the Yellow River Basin (YRB), China’s vital ecological security barrier and economic belt. However, existing research lacks a spatial perspective on the coordinated development between TI and inclusive green growth (IGG), with limited understanding of cross-regional spillover mechanisms. Based on panel data from 75 cities in the YRB (2011–2023), this study constructs a comprehensive evaluation system encompassing the scale, structure, and potential dimensions of the TI and the economic, social, livelihood, and environmental dimensions of IGG. The study employs the coupling coordination degree (CCD) model, exploratory spatial data analysis (ESDA), and the Spatial Durbin Model (SDM) to examine spatiotemporal evolution and spillover effects. The results reveal an upward yet fluctuating coordination trend with pronounced spatial heterogeneity, characterized by a “downstream–midstream–upstream” gradient pattern, dual-core radiation centered on the Jinan–Qingdao and Xi’an–Zhengzhou agglomerations, and persistent High–High clusters in the Shandong Peninsula contrasted with Low–Low clusters in the upstream Qinghai–Gansu–Ningxia region. Critically, new-quality productive forces exert significant positive direct and spillover effects, while industrial structure and government intervention have inhibitory spatial effects on adjacent cities. Regional heterogeneity analysis confirms factor-endowment-driven differentiation across upstream, midstream, and downstream areas. These findings advance spatial spillover theory in river basin contexts and provide evidence-based pathways for balancing economic growth with ecological protection in ecologically sensitive regions worldwide, directly supporting multiple UN Sustainable Development Goals.

1. Introduction

As a crucial ecological security barrier and core belt of regional economic development [1], the high-quality development of the Yellow River Basin (YRB) is not only related to the overall situation of regional coordinated development but also has strategic significance for national ecological security and national sustainable development. In the year 2019, ecological conservation paired with high-quality development within the YRB was upgraded to the status of a national strategic priority [2], with a clear proposal to “explore a new path of high-quality development with regional characteristics.” This strategic positioning provides fundamental guidance and directional reference for the multi-dimensional coordinated development of economic upgrading, ecological protection, and social progress within the basin. As a highly potential strategic pillar industry in the YRB, the tourism industry (TI) serves as an economic driving force featuring ecological friendliness and social inclusiveness [3]. It can not only create significant economic value by relying on the basin’s unique natural and cultural resources but also ensure the protection and restoration of the ecological environment through ecotourism development [3]. Furthermore, it can provide diversified employment opportunities for residents along the basin and narrow the urban–rural income gap, thus serving as a key link and practical vehicle that accurately connects ecological protection with economic and social development. However, tourism development within the basin exhibits pronounced spatial imbalance. In 2023, Shandong Province in the downstream region received 800 million tourists and generated CNY 910 billion in tourism revenue, whereas Qinghai Province in the upstream region received only 44.76 million tourists, with a revenue of CNY 43.06 billion—differences of approximately 21-fold and 18-fold, respectively. This “downstream-prosperous, upstream-lagging” pattern reflects the broader developmental gradient where upstream provinces consistently rank among the lowest nationally, highlighting the urgent necessity for research on coordinated tourism development within the YRB. By comparison, inclusive green growth (IGG) serves as an emerging development paradigm that balances economic growth, ecological conservation, and social equity simultaneously [4] and emphasizes reducing environmental pressure while ensuring that the fruits of growth benefit all residents. It is highly consistent with the strategic requirements of the YRB, such as “ecology first, green development” and “common prosperity.” As such, exploring the coordinated development relationship between TI and IGG is not only an inevitable choice to proactively respond to major national strategic needs but also a key path to address practical dilemmas in the basin, including the superposition of ecological fragility and economic underdevelopment [1], as well as unbalanced and inadequate development.

The concept of IGG was first proposed by the World Bank in 2012. In essence, it is a new sustainable development paradigm that balances environmental sustainability and social equity. There exists an inherent relationship of mutual empowerment and in-depth correlation between the TI and IGG [5]. The literature closely related to the research theme of this paper mainly focuses on the following three core directions.

(1) The mechanism of the relationship between the TI and IGG: Based on the concept of IGG, scholars have conducted research around the inherent logic and practical strategies of the TI development. Li took the lead in pointing out that the industrial attributes and characteristics of the TI are naturally consistent with the connotation of inclusive growth [6]. Wang further introduced the concept of inclusive growth into the field of tourism research, systematically exploring the mechanism through which inclusive tourism growth can be realized [7]. Li et al. found that the concept of inclusive growth can effectively resolve conflicts of interest among subjects such as governments, enterprises, and community residents in tourism development [8]. Regarding the impact of the TI on IGG, Guo and Lin proposed that rural tourism and tourism in ethnic minority areas are effective paths to achieving local inclusive growth, respectively [9,10]. Jeyacheya et al. pointed out, based on data from developing countries, that the increase in employment opportunities, growth in residents’ income, and improvement in tax effects brought by the TI can significantly enhance inclusive growth in the short term [11]. Zhang and Chen found through empirical research that tourism development has a promotive effect on inclusive growth in contiguous destitute areas, but this effect shows obvious urban–rural differentiation characteristics after the improvement of transportation conditions [12]. In addition, based on the coupling coordination theory, Wang et al. carried out a quantitative examination of the coupling coordination interaction between tourism development and IGG in Hunan Province as well as their influencing elements, and the study verified that marked discrepancies exist in the coupling coordination degree (CCD) between these two dimensions with respect to temporal change and spatial patterns [5].

(2) Measurement methods and development models of the TI’s inclusive green growth: In the field of measurement research, scholars have constructed distinctive evaluation index systems based on different research perspectives and analytical frameworks. Zhong et al. focused on the core dimensions of inclusive growth, built an evaluation index system for the TI’s inclusive growth from three aspects, namely, fairness, effectiveness, and shareability; adopted the improved entropy weight method to quantitatively measure the inclusive growth status of the TI in various provinces of China; and further compared the differences between the measurement results and total tourism income, providing a reference for understanding the matching relationship between the “quality” of inclusive growth and the “scale” of tourism economy [13]. Li continued the research on inclusive growth measurement, adopted a process-oriented analytical logic, constructed an evaluation system from three progressive fields—”preconditions—growth process—growth results”, and conducted a targeted evaluation of the inclusive growth level of the TI in Shaanxi Province, improving the regional adaptability of measurement research [14]. At the same time, the exploration of the TI’s green growth measurement has gradually advanced. To fill the gap in green growth evaluation, Ming drew on the core elements of the OECD green growth conceptual framework, combined with the general process of tourism production activities, and systematically designed the first evaluation index system for the TI’s green growth, laying a foundation for quantitative research in this field [15]. Tian et al. integrated the United Nations Environment Programme (UNEP) inclusive green economy measurement framework with the development characteristics of China’s tourism industry, optimized and constructed a green growth evaluation system for the TI from the perspective of industrial growth, calculated China’s TI green growth index using the entropy weight-TOPSIS method, and deeply analyzed its evolution process, enriching the methods and conclusions of green growth measurement [16]. In terms of model summary, Wang and Luo systematically summarized the practical model of inclusive growth of Thailand’s tourism economy from four key dimensions—industrial balance, equal opportunities, fiscal inclusion, and rights protection—by sorting out international experiences, providing a reference for other countries and regions [17]. Based on the practices of specific regions in China, Wang and Wang further proposed a tourism inclusive development model for China’s ethnic minority areas. With tourism economic development as the core foundation, this model also attaches great importance to the social development of tourism communities, aiming to achieve the inclusive development goal of “dual-track parallel” economy and society and providing path references for regional tourism development with Chinese characteristics [18].

(3) Identification of the factors influencing the TI’s inclusive green growth: Tian et al. verified via regression analysis that economic development level and tourism industry structure constitute the core determinants influencing the IGG driven by the TI [16]. Song and Liu focused on the differentiated role of the unbalanced development of the digital economy on the TI’s inclusive green growth [19]. Liu et al. systematically tested the impact mechanism of social capital on the TI’s inclusive green growth using models such as threshold effects [20]. Liu et al. employed approaches including dynamic panel models for the empirical examination of the impacts and regional heterogeneity characteristics of digital infrastructure development on IGG driven by the TI while also conducting an in-depth exploration of its underlying action mechanisms [21].

Undoubtedly, prior scholarly work has established a robust groundwork for advancing investigations into the nexus between TI and IGG; however, there remains considerable room for further development: First, existing studies have mainly focused on national, provincial, or single-city units, with insufficient attention paid to the coordinated development of the TI and IGG in the YRB—a crucial national ecological barrier and economic belt where the economic–social-environmental system is undergoing rapid transformation. Second, most existing studies are confined to exploring the coordinated development of the TI and IGG within a single region, failing to fully consider cross-regional coordinated development. Few studies have addressed how to achieve a higher level of coordinated development between TI and IGG from the perspective of spatial correlation, and the exploration of the inherent mechanisms and spatiotemporal evolution patterns of their coordinated development remains insufficient. Third, in revealing the influencing factors of the coordinated development of the TI and IGG, previous studies have often been conducted based on the assumption of spatial homogeneity, ignoring the important role of spatial interaction effects, which makes it difficult to interpret the spatial spillover effects and their inherent influencing mechanisms of the coordinated development between the two. To sum up, this study selects 75 prefecture-level cities (prefectures, leagues) across the YRB during the period 2011–2023 as research subjects, systematically establishes a mechanistic framework and a comprehensive evaluation indicator system for the coordinated development between TI and IGG, examines the spatiotemporal evolutionary traits of their coordinated development across the YRB by adopting methodologies including the entropy weight method, coupling coordination degree model, and ESDA, and ultimately identifies their spatial spillover impacts through the application of the Spatial Durbin Model (SDM). In comparison with prior research, the marginal contributions of this study are manifested in the following respects: (1) Taking the YRB as the research sample and targeting its regional characteristics as a national ecological security barrier and a core belt for high-quality development, this paper systematically explores the coordinated relationship between TI and IGG from both theoretical and empirical perspectives. The research is more in line with national strategic needs, and the conclusions are more regionally targeted. (2) From the dimension of spatial correlation, this study emphasizes elucidating the multi-dimensional spatiotemporal evolutionary traits of the coordinated development between TI and IGG—a focus that serves as a meaningful supplement to prior scholarly work. (3) It systematically analyzes the spatial spillover effects and their influencing mechanisms of the coordinated development of the TI and IGG in the YRB, providing a new explanation for examining the drivers of spatial spillover effects and making the research conclusions more reliable. Notably, the analytical logic and empirical conclusions of this paper can be further aligned with core goals of the United Nations Sustainable Development Goals (SDGs), such as SDG1, SDG3, SDG4, SDG8, and SDG13, providing a replicable research paradigm for similar river basins and ecologically sensitive areas worldwide. On the one hand, the coordinated development model of the TI and IGG in the YRB can be extended to regions with similar ecological endowments and development demands. By adjusting index weights and adapting to regional characteristics, it can provide references for formulating differentiated coordinated paths. On the other hand, the research conclusions on spatial spillover effects can provide references for the formulation of cross-regional ecological compensation and industrial linkage policies, helping other regions balance ecological protection and social equity while achieving economic growth, and ultimately promoting the localized implementation and effective practice of SDGs at the regional level.

2. Theoretical Framework and Mechanism Analysis

2.1. Coupling Coordination Theory

Based on the coupling coordination theory [22], the coordinated development of the TI and IGG essentially reflects the objective dynamic interaction relationship between the elements of these two regional systems. Its ultimate goal is to promote the region’s realization of the ecology-first, green development and common prosperity, forming a development pattern of mutual empowerment and coordinated progress.

2.2. Interaction Mechanisms Between TI and IGG

2.2.1. TI’s Driving Effects on IGG

TI provides new momentum for IGG. TI exerts its influence primarily through industrial scale expansion, industrial structure optimization, and the release of development potential: (1) Scale-driven effect. Being a typically labor-intensive sector, the scaled expansion of the TI is capable of generating substantial employment opportunities—a phenomenon that contributes to narrowing the urban–rural income disparity and thus facilitates the attainment of social inclusion objectives [11]. Meanwhile, the increase in total tourism income can be reinvested into ecological protection, strengthen the economic driving force of IGG, and consolidate its ecological constraint bottom line. (2) Structure-optimizing effect. Industrial structure optimization gives rise to high-value-added new tourism formats such as cultural-tourism integration and agriculture-tourism integration, enriches the supply of tourism products, realizes the fair distribution of tourism income, and promotes the upgrading of the social inclusion and livelihood development dimensions of IGG [9]. (3) Potential-leading effect. Relying on emerging technologies such as tourism big data and artificial intelligence, TI releases its development potential. Its technological empowerment role can accurately match the tourism supply-demand relationship, reduce resource waste and environmental consumption, and provide technical support for the sustainability of IGG [23]. Furthermore, the cultivation of new business models such as tourism cooperatives and community-participatory tourism further strengthens the equity attribute of the social inclusion dimension of IGG [12,24].

2.2.2. IGG’s Supporting Effects on TI

IGG provides reverse support and path guidance for the development of the TI by establishing a clear development direction. Guided by the concept of coordinating economy, environment, society, and people’s livelihood, IGG aims to realize the integrated advancement of economic growth, social equity, and ecological protection. IGG mainly exerts the following effects: (1) Economic support effect. As the material guarantee for the industrial scale expansion of the TI, the economic growth dimension of IGG not only releases tourism consumption potential by improving residents’ income levels but also injects market momentum into the development of the TI, optimizes the investment structure of the TI, promotes the construction of infrastructure such as transportation networks and distribution centers, alleviates the spatiotemporal constraints on the cross-regional flow of tourism factors, and ultimately consolidates the hardware foundation for the large-scale development of the TI [25]. (2) Social escort effect. The social inclusion dimension of IGG provides a fair institutional environment for the industrial structure optimization of the TI. The narrowing of the urban–rural development gap promotes the transformation of the rural tourism market from a potential form to an actual form, driving the adjustment of the tourism industrial structure to improve product supply [26]. The equalization of public services improves the conditions for human capital accumulation such as education and skill training of rural residents, provides high-quality labor support for TI, and thereby achieves a substantial improvement in tourism service quality. (3) Livelihood-driven effect. The livelihood development dimension of IGG provides demand momentum for the release of the TI’s development potential. With the increase in residents’ demand for a better life, the demand for high-quality tourism consumption such as cultural experience, health and wellness leisure, and ecological sightseeing has grown rapidly, driving TI to transform towards meeting people’s livelihood and well-being needs in format innovation, product supply, and service optimization. (4) Ecological regulation effect. The green environmental protection dimension of IGG provides framework constraints and direction guidance for the sustainable development of the TI. Policy tools such as the delineation of ecological protection red lines and the control of scenic area environmental capacity promote the green development of the TI. The implementation of green innovation application scenarios such as new energy application and smart energy management can effectively reduce ecological degradation in the process of tourism development and improve tourism green efficiency and competitiveness [27].

2.3. Spatial Spillover Effects of Coordinated Development

The coordinated development of the TI and IGG does not exist in isolation, and its level evolution is often significantly affected by adjacent regions. Based on the new economic geography theory [28], as a key path for regional high-quality development, the coordinated development of the two possesses spatial spillover characteristics driven by the cross-border flow of factors. Affected by regional development gradient differences, the spatial spillover effect is mainly manifested as a dynamic game between the siphon effect and the diffusion effect. Drawing on the regional unbalanced development theory [29], the formation of the siphon effect stems from the fact that regions with a high level of coordinated development attract high-quality factors from low-level regions by virtue of comprehensive advantages such as technological innovation, policy efficiency, and openness, thereby widening the regional development gap—a mechanism that has also been verified by the research of Cai et al. [30]. In contrast, the diffusion effect can strengthen cooperation in the coordinated development of the TI and IGG among regions, helping low-level regions imitate, learn from, and draw on the advanced experience of high-level regions in technology application, policy regulation, and industrial operation. This in turn drives low-level regions to improve factor allocation, optimize development paths, move towards a higher coordination level, and promote the balanced development of the region as a whole [31]. Although regional integration strategies strengthen the linkages among comprehensive factors among regions, the optimal flow of factors in low-level regions in the early stage makes the siphon effect dominant and the spillover effect negative. With the deepening of integration, the “crowding effect” caused by factor saturation in high-level regions will promote factor reflux, making the diffusion effect dominant and the spillover effect turning positive. In addition, if low-level regions are constrained by traditional path dependence or unbalanced factor allocation, there may be no significant spatial spillover in the coordinated development of the two [32].

3. Materials and Methods

3.1. Study Area

This paper focuses on the YRB, a trans-regional basin spanning China’s eastern, central, and western regions [33]. Geographically, it covers nine provinces and autonomous regions, namely Qinghai, Sichuan, Gansu, Inner Mongolia, Ningxia, Shanxi, Shaanxi, Henan, and Shandong, with a basin area of 795,000 square kilometers (including 42,000 km² of internal drainage area) [33]. As a critical ecological barrier and core economic belt in China [1], the YRB undertakes pivotal ecological functions, including water conservation, soil erosion control, and biodiversity protection [33]. In September 2019, the ecological protection and high-quality development of the YRB was elevated to a major national strategy [2], highlighting its strategic significance in China’s regional development layout.

The YRB represents an ideal case for investigating the coordinated development between TI and IGG for several compelling reasons. First, the basin exemplifies the typical superposition of ecological fragility and economic underdevelopment [1], making it a critical testbed for exploring pathways that balance environmental protection with economic growth. Second, the YRB possesses abundant and distinctive tourism resources, including the Yellow River culture—the cradle of Chinese civilization—along with unique natural landscapes such as the Loess Plateau, wetland ecosystems, and diverse historical and cultural heritage sites [34]. These resources provide substantial potential for tourism-driven inclusive green development. Third, the pronounced development gradient across upstream, midstream, and downstream regions [33] offers an ideal spatial laboratory for examining heterogeneous coordination patterns and cross-regional spillover effects. Fourth, as a nationally strategic region, research findings from the YRB can directly inform policy formulation for ecological protection and high-quality development [34]. Finally, the analytical framework and empirical conclusions derived from the YRB can be extended to similar ecologically sensitive river basins worldwide, contributing to the localized implementation of the United Nations Sustainable Development Goals (SDGs).

This paper takes the natural river basin scope delineated by the Yellow River Conservancy Commission of the Ministry of Water Resources as its research boundary [33], selecting 75 prefecture-level cities within the basin as research objects and excluding non-YRB core cities (e.g., Hanzhong, Hulunbuir, which belong to other economic zones) and cities with significant data scarcity (e.g., Haidong, Qinghai) (Figure 1).

3.2. Construction of the Index System

Given the comprehensive development attributes of the TI, single indicators such as total tourism income or total tourist arrivals are obviously unable to comprehensively reflect core development dimensions of the TI, including enterprise operational efficiency, employment absorption quality, and market expansion breadth. Therefore, this paper resorts to a multi-dimensional comprehensive evaluation index system to quantify the development level of urban TI. Regarding the measurement methods of IGG, the comprehensive index method and efficiency measurement method are currently the most widely used in academic circles. Among them, although the efficiency measurement method can accurately measure the level of IGG efficiency by calculating the relative distance between the technology frontier and the technical level of Decision-Making Units (DMUs) [35], its core logic lies in assessing the ability of DMUs to convert input factors into output factors. This is essentially different from the development concept of coordinating and balancing the economy, environment, society, and people’s livelihood emphasized by IGG in this paper. Thus, this paper ultimately adopts the comprehensive index method to conduct a comprehensive evaluation of IGG in various cities of the YRB. Based on the conceptual connotations and complex characteristics of the TI and IGG, and with reference to the existing evaluation frameworks for TI development competitiveness and IGG [36,37,38], this paper further integrates the core goals of “inclusiveness” and “sustainability” in the United Nations’ 2030 Agenda for Sustainable Development. Meanwhile, taking into account the comprehensiveness, scientificity, representativeness, and data availability of the index system, a comprehensive evaluation index system including the TI subsystem and IGG subsystem is constructed (

Table 1. Comprehensive Evaluation Index System of TI and IGG.

Sub-System	Primary Indicators	Secondary Indicators	Unit	SDGs
TI	Industrial Scale	Domestic tourism reception person-trips	10,000 person-trips	SDG8
		Inbound tourism reception person-trips	10,000 person-trips	SDG8
		Share of domestic tourism revenue in GDP	%	SDG8
		Share of tourism foreign exchange revenue in GDP	%	SDG8
	Industrial Structure	Number of star-rated hotels	Hotels	SDG8
		Number of scenic areas above 4A level	Scenic areas	SDG8
		Fixed asset investment	10,000 yuan	SDG9
	Development Potential	Employment in the tertiary industry	10,000 people	SDG8
		Resident consumption level	Yuan	SDG12
		Share of total tourism revenue in GDP	%	SDG8
IGG	Economic Growth	Per capita GDP	Yuan	SDG8
		GDP growth rate	%	SDG8
		Share of tertiary industry in GDP	%	SDG8
		Average wage of employees	Yuan	SDG8
		Urbanization rate	%	SDG8
	Social Inclusion	Per capita disposable income of urban residents	Yuan	SDG1
		Per capita disposable income of rural residents	Yuan	SDG1
		Ratio of per capita disposable income of urban residents to rural residents		SDG1
	Livelihood Development	Participation rate of basic endowment insurance for urban employees	%	SDG1
		Number of schools per 10,000 people	Schools per 10,000 people	SDG4
		Number of teachers per 10,000 people	Teachers per 10,000 people	SDG4
		Number of books in public libraries per 10,000 people	Books per 10,000 people	SDG4
		Number of hospitals per 10,000 people	Hospitals per 10,000 people	SDG3
		Number of hospital beds per 10,000 people	Beds per 10,000 people	SDG3
	Green Environmental Protection	Greening coverage rate of built-up areas	%	SDG11
		Industrial Nitrogen Oxide Emissions per unit of GDP	Tons per 10,000 yuan	SDG13
		Industrial wastewater emissions per unit of GDP	Tons per 10,000 yuan	SDG13
		Industrial smoke and dust emissions per unit of GDP	Tons per 10,000 yuan	SDG13
		Comprehensive utilization rate of general industrial solid waste	Tons per 10,000 yuan	SDG12
		Centralized treatment rate of sewage treatment plants	%	SDG12
		Harmless treatment rate of municipal solid waste	%	SDG12

Source: Authors’ compilation. Data from China City Statistical Yearbook and Provincial Statistical Yearbooks (2011–2023).

). Specifically, the TI subsystem consists of 3 primary indicators: industrial scale (reflecting development foundation), industrial structure (embodying development quality), and development potential (characterizing growth momentum), with 10 secondary indicators under them. It fully covers the three-dimensional development characteristics of the TI, namely, “scale-quality-potential”. The IGG subsystem includes 4 primary indicators: economic growth (development driver), social inclusion (equity dimension), livelihood development (welfare goal), and green environmental protection (ecological constraint), with 21 secondary indicators. Among these, economic growth and green environmental protection are the objective quantification of the development economics concept of “synergy between economic growth and ecological sustainability”, while social inclusion and livelihood development centrally reflect the core demands of welfare economics, namely “intragenerational equity and intergenerational inheritance”, which fully align with the theoretical connotation of IGG as “multi-dimensional balance”.

Relevant data are mainly derived from the China City Statistical Yearbook, Statistical Yearbooks of various prefecture-level cities in the YRB, and Statistical Communiqués on National Economic and Social Development of those cities. For a limited number of missing values in individual years—primarily affecting indicators such as inbound tourism reception person-trips and industrial pollutant emissions per unit of GDP in certain upstream cities where statistical infrastructure is relatively underdeveloped—the linear interpolation method was employed to estimate values based on adjacent observed data points [39]. This approach assumes a linear trend between consecutive periods and is widely adopted in panel data studies when missing values are sporadic and the underlying trend is relatively stable [40]. Similar applications of linear interpolation for handling missing data in Chinese regional tourism and environmental research have been validated in recent studies [41]. The proportion of interpolated data remains below 5% of the total dataset, ensuring that the interpolation does not substantially affect the reliability of the empirical results.

3.3. Research Methods

3.3.1. Entropy Weight Method

The entropy weight method is an objective weighting technique that determines indicator weights based on the degree of data dispersion, effectively avoiding subjective bias inherent in expert-based methods such as the Analytic Hierarchy Process. This approach has been widely adopted in regional development and tourism research for multi-indicator comprehensive evaluation. Compared with alternative weighting approaches, the entropy weight method offers distinct advantages for this study: (1) Unlike expert-assigned weights, which require subjective judgment and may be inconsistent when assessing 75 cities across 13 years with substantial regional heterogeneity—from ecological indicators in upstream areas to economic indicators in downstream regions—the entropy weight method automatically adapts to this diversity through data-driven weight allocation. (2) Unlike PCA-based weights, which extract principal components with relatively weak explanatory power for original indicators, the entropy weight method directly assigns weights to each of the 31 secondary indicators, providing clearer interpretation of their individual contributions. (3) Unlike equal weighting, which assumes all indicators are equally important and may dilute critical information from high-dispersion indicators, the entropy weight method appropriately captures differential information contributions across economic, ecological, and social dimensions. The entropy weight method can measure the contribution of the degree of data dispersion in the system to the comprehensive evaluation and overcome the impacts of information overlap among multiple indicator variables and subjectivity on weight determination. This paper adopts the entropy weight method incorporating time variables to determine the weights of each indicator; for specific calculation formulas, refer to Yang et al. [42].

3.3.2. CCD Model

The coupling coordination degree (CCD) model, originating from the coupling concept in physics, is designed to quantify the degree of mutual influence and synergistic development between two systems [43]. Compared with unidirectional analytical methods such as regression analysis, the CCD model can effectively capture bidirectional interactions and has been extensively applied in tourism-environment and urbanization-ecology coordination studies [44]. This paper applies the CCD model to evaluate the coordination relationship between the TI and IGG. The formula is as follows:

$$D=\sqrt{C\times T}$$

(1)

$$T=\alpha W_{1m}+\beta W_{2m}$$

(2)

where $D$ denotes the CCD between the TI and IGG, ranging from 0 to 1; $C$ represents the coupling degree between TI and IGG; $T$ denotes the comprehensive coordination index; $W_{1m}$ denotes the tourism industry index score of city m; $W_{2m}$ denotes the inclusive green growth index score of city m; $\alpha$ and $\beta$ are undetermined coefficients, with their sum equal to 1. Considering that TI and IGG are equally important, this paper sets both $\alpha$ and $\beta$ to 0.5. Drawing on existing studies [45], 10 CCD stages are divided (

Table 2. Classification standards of CCD.

CCD Range	Coordination Grade
(0.00, 0.10]	Extremely imbalance
(0.10, 0.20]	Severely imbalance
(0.20, 0.30]	Moderately imbalance
(0.30, 0.40]	Mildly imbalance
(0.40, 0.50]	Borderline imbalance
(0.50, 0.60]	Barely coordinated
(0.60, 0.70]	Primary coordination
(0.70, 0.80]	Intermediate coordination
(0.80, 0.90]	Good coordination
(0.90, 1.00]	High-quality coordination

Source: Adapted from Zhang et al. [45].

):

3.3.3. ESDA

For the purpose of further examining the spatial correlation characteristics and disparity levels of the coordinated development between TI and IGG in the YRB, this study adopts the ESDA methodology. The Global Moran’s I index is utilized to measure the overall spatial correlation and spatial disparity of the research subjects—a tool that enables effective identification of the holistic spatial traits and spatial disparities pertaining to the coordinated development between TI and IGG. In contrast, the Local Moran’s I index is utilized to explore the evolutionary pattern of the coordinated development between TI and IGG as well as the agglomeration distribution of outliers.

$$I={\displaystyle \frac{{\sum }_{i=1}^n{\sum }_{j\ne i}^n{W}_{ij}{Z}_i{Z}_j}{{\sigma }^2\sum _{i=1}^n\sum _{j\ne i}^n{W}_{ij}}},\left[Z_i={\displaystyle \frac{V_i-\bar{V}}{\sigma }}, \bar{V}={\displaystyle \frac{1}{n}}\sum _{i=1}^n{V}_i, \sigma ={\displaystyle \frac{1}{n}}\sum _{i=1}^n{\left(V_i-\bar{V}\right)}^2\right]$$

(3)

$$\mathrm{Local}{\mathrm{Moran}}^{’}\mathrm{s}I=Z_i\sum _{i=1}^n{W}_{\mathrm{ij}}{Z}_j$$

(4)

where I denotes the Global Moran’s I index; n represents the sample size; Z_i is the standardized transformation of $V_i$; $W_{ij}$ denotes the spatial weight matrix, which takes a value of 1 if city i is adjacent to city j, and 0 otherwise. The Global Moran’s I index ranges between −1 and 1: a positive value of I (I > 0) indicates spatial positive correlation; a negative value (I < 0) indicates spatial negative correlation; and a value of 0 (I = 0) indicates random spatial distribution. A positive Local Moran’s I value signifies the spatial agglomeration of elements with the same type of attribute values, while a negative value indicates the spatial agglomeration of elements with different types of attribute values.

3.3.4. SDM

The SDM fully incorporates the spatial autocorrelation of both explanatory and explained variables, thereby facilitating more precise identification of the spatial autocorrelation inherent in the coordinated development between TI and IGG within the YRB, along with the extent of the impacts exerted by diverse influencing factors on such coordinated development. The model specification is as follows:

Where $Y_{it}$ and $X_{it}$ represent the observed values of the explained variable and explanatory variables for the i-th research unit in period t, respectively; $\beta$ denotes the coefficient of explanatory variables; ρ is the spatial autoregressive coefficient of the explained variable; $\phi$ represents the spatial spillover coefficient; $W_{ij}$ denotes the spatial weight matrix; ${\mu }_i$, ${\nu }_t$, and ${\epsilon }_{it}$ denote the spatial effects, time effects, and random error terms, respectively.

$$Y_{it}=\beta X_{it}+\rho \sum _{j=1}^n{W}_{ij}{Y}_{jt}+\sum _{j=1}^n\phi W_{ij}{X}_{jt}+{\mu }_i+{\nu }_t+{\epsilon }_{it}$$

(5)

4. Results

4.1. Temporal Evolutionary Characteristics

The temporal variation trend of the CCD between TI and IGG in the YRB over the period 2011–2023 is depicted in Figure 2. It reveals that during the sample period, the CCD displayed a fluctuating pattern alongside an overall upward trend, with the level of coordinated development continuously improving. This suggests that the coordinated development of these two systems within the YRB underwent a gradual enhancement over the study period. Based on the interval of the CCD, it can be roughly divided into two stages: (1) The slowly rising stage (2011–2019). The average annual growth rate of the CCD was approximately 1.58%, and the coordinated development level gradually approached primary coordination from barely coordinated. Since the 18th National Congress of the Communist Party of China, the implementation of a series of national policies has brought opportunities for the development of the YRB. The YRB has actively responded to the ecological protection and high-quality development strategy. In terms of the TI, it has increased the development of ecotourism resources and promoted the integrated development of tourism with other industries; in terms of green growth, it has strengthened environmental governance and ecological restoration and improved resource utilization efficiency. These measures have promoted the continuous improvement of the development levels of the TI and IGG, and the CCD between them has also steadily increased. (2) The fluctuating adjustment stage (2020–2023). Affected by factors such as the sudden public health emergency, the CCD fluctuated, with an average annual growth rate of approximately −0.47%. In 2020, the CCD dropped to 0.532, and although it recovered slightly thereafter, it remained generally lower than the 2019 level. As a labor mobility-intensive industry, TI suffered a severe impact during this period, with frequent occurrences such as shrinkage of the tourism market and restrictions on tourism projects. This hindered the coordinated development of the TI and IGG and affected the process of improving the CCD.

4.2. Spatial Distribution Characteristics

Figure 3 depicts the spatial distribution pattern of synergistic development between TI and IGG within the YRB across the years 2011, 2015, 2019, and 2023. It can be observed that: (1) The coordinated development generally presents a spatial differentiation characteristic of being high in the downstream, medium in the midstream, and low in the upstream. The Shandong Peninsula in the downstream (e.g., Qingdao, Jinan) has long maintained a level of primary coordination or above, relying on its mature tourism industry system and green development foundation; the Henan-Shanxi region in the midstream (e.g., Zhengzhou, Xi’an) is mostly in the transition from barely coordinated to primary coordination; the Qinghai-Gansu-Ningxia-Inner Mongolia region in the upstream is dominated by mildly imbalanced and borderline imbalanced. This pattern is directly related to the spatial matching of the basin’s economic development level, tourism resource endowments, and ecological constraint intensity. The downstream boasts strong economic vitality and a mature tourism market; the midstream possesses both cultural and ecological resources but faces significant industrial transformation pressure; the upstream is ecologically sensitive yet lags behind in tourism development, forming a stepped differentiation. (2) Cities with high-level coordinated development have gradually formed dual-core radiation belts centered on the Jinan-Qingdao Urban Agglomeration and the Xi’an-Zhengzhou Urban Agglomeration in space. In 2011, the dual-core agglomerations existed merely in a scattered distribution; by the year 2019, the count of cities achieving good coordination within the Qingdao Urban Agglomeration had increased to 5, while the number of cities with intermediate coordination in the Xi’an-Zhengzhou Urban Agglomeration had hit 7. Although there was a slight decline due to external shocks in 2023, the core areas still maintained a level above primary coordination. The Qingdao Urban Agglomeration has formed synergy relying on marine tourism and green port economy, while the Xi’an-Zhengzhou Urban Agglomeration has strengthened linkages through the cultural and tourism brand of “the starting point of the Silk Road” and low-carbon industrial policies, thereby radiating to the surrounding areas. (3) The coordinated development shows an unbalanced “core-periphery” distribution, with limited diffusion from core cities to peripheral cities. From 2011 to 2019, the radiation effect from core areas to peripheral areas was significant (e.g., Jinan’s driving effect on Liaocheng, Xi’an’s driving effect on Xianyang), but most peripheral cities (e.g., Longnan, Wuzhong) were still in a state of borderline imbalanced. Affected by external shocks in 2020, both core and peripheral areas experienced a decline in coordination levels. However, the core areas recovered rapidly in 2023, while the peripheral areas achieved a leap from mildly imbalanced to barely coordinated due to policy support, reflecting the resilience of core areas and the policy-driven growth characteristics of peripheral areas.

4.3. Spatial Agglomeration Characteristics

The results of the global Moran’s I index calculation for the CCD between TI and IGG in the YRB from 2011 to 2023 are presented in

Table 3. Global Moran’s I and test results of the coordinated development between TI and IGG in the YRB from 2011 to 2023.

Year	Moran’s I	Z-Scores	p-Value
2011	0.469	6.171	0.000
2012	0.467	6.132	0.000
2013	0.480	6.298	0.000
2014	0.423	5.571	0.000
2015	0.376	4.972	0.000
2016	0.391	5.516	0.000
2017	0.391	5.166	0.000
2018	0.380	5.031	0.000
2019	0.362	4.793	0.000
2020	0.243	3.296	0.001
2021	0.219	2.624	0.009
2022	0.262	3.544	0.000
2023	0.227	3.081	0.002

: (1) All Moran’s I indices are significantly positive, with values ranging from 0.219 to 0.480. Furthermore, the p-values for all years are less than 0.01 and the Z-values are greater than 2.624, passing the rigorous statistical significance test. This indicates that the coordinated development of the TI and IGG in the YRB is not randomly distributed but exhibits significant positive spatial autocorrelation. Specifically, cities with high coordination levels have formed strong alliances due to their adjacency, while cities with low coordination levels have also clustered together. In essence, this spatial agglomeration characteristic is a manifestation of the spatial dependence effect in the YRB—the convenience of cross-regional tourism factor flow, the cross-administrative boundary relevance of the ecological environment, and the spatial spillover effect of policy implementation have collectively constructed a spatial linkage network of coordinated development characterized by mutual prosperity and mutual decline. (2) Over the study period, the Moran’s I index fluctuated and decreased from 0.469 in 2011 to 0.227 in 2023, indicating that the spatial agglomeration characteristics of the coordinated development between TI and IGG in the YRB have weakened with fluctuations over time, showing a convergence trend.

To further reveal the clustering or diffusion characteristics of the coordinated development between TI and IGG in the YRB, the years 2011, 2015, 2019, and 2023 were selected as temporal cross-sections to draw LISA clustering maps (Figure 4). The findings indicate the following: (1) The spatial clustering of coupling coordination development in the YRB presents the characteristic of “the strong become stronger while the weak become weaker”, dominated by High-High (HH) and Low-Low (LL) clusters, with High-Low (HL) and Low-High (LH) clusters as supplements. Although the proportion of cities with HH and LL clusters has shown a slow decline, they have always occupied an absolute dominant position, confirming the “polarization-lock-in effect” of the basin’s coordinated development: cities with high coordination degrees form synergistic clusters through tourism resource sharing and interconnected ecological governance, while cities with low coordination degrees fall into “low-level lock-in” due to weak development foundations and insufficient policy support. (2) In terms of spatial distribution, HH clusters have long been locked in the Shandong Peninsula in the downstream region (e.g., Qingdao, Yantai, Weihai). Relying on coastal tourism resources and green industrial foundations, these cities have formed “high-value clusters” in coordination degrees; LL clusters are mainly distributed in the Qinghai-Gansu-Ningxia-Inner Mongolia region in the upstream, showing the characteristic of “low-value clusters” due to ecological sensitivity constraints and lagging tourism development. This distribution is highly consistent with the basin’s development gradient of “active economy in the downstream and ecological constraints in the upstream”, reflecting the spatial matching between tourism resource endowments and ecological policy constraints. (3) In 2011, HH clustering cities were concentrated in the eastern part of the Shandong Peninsula; by 2019, they had spread to inland cities such as Jinan and Zibo; in 2023, they shrank to core cities such as Qingdao and Yantai. This is related to the diversion of resources caused by the basin’s tourism policies tilting towards the midstream, such as the construction of the “Yellow River Cultural Tourism Belt”. In 2011, LL clusters were continuously distributed in the upstream; in 2023, they differentiated into isolated low-value points such as Xining and Wuzhong. Some cities such as Lanzhou broke away from the LL cluster driven by new-quality productive forces, showing the role of policy intervention and technological progress in breaking low-value lock-in. The HL and LH clusters are extremely small in number and scattered in distribution, serving as transition zones between “high-value” and “low-value” regions. However, due to the lack of policy coordination and factor flow, they have failed to form effective radiation, reflecting the “spatial disconnection” of the basin’s spatial correlation.

4.4. Spatial Spillover Effects Analysis

4.4.1. Variable Selection

(1)

Explained variable. This paper selects the CCD between TI and IGG in the YRB as the explained variable.

(2)

Explanatory variables. Drawing on relevant scholars’ research, the following variables are chosen as explanatory variables for the empirical model:

Industrial structure (struc): Measured by the proportion of the added value of the secondary industry in GDP [46].

Financial development level (finance): Measured by the proportion of the balance of deposits and loans of financial institutions at the end of the year in GDP [47].

Government intervention degree (gov): Measured by the proportion of local general public budget expenditure in GDP [48].

New-quality productive forces (npro): Drawing on the research of Han et al. [49]. and Lu et al. [50], it is quantitatively characterized based on three dimensions—new-quality labor force, new-quality labor objects, and new-quality means of labor—and the entropy weight method is adopted to objectively evaluate the new-quality productive forces level of each city.

Opening-up degree (open): Measured by the proportion of total import and export of goods in GDP [51].

4.4.2. Spatial Econometric Model Test and Selection

Variable stationarity is a prerequisite for panel regression estimation. Based on the results of the LLC test (for common roots) and Fisher-ADF test (for individual roots), all variables are stationary. The Kao test is employed to conduct a cointegration test on the regression model, and the results indicate a significant long-term stable equilibrium relationship between the CCD and various influencing factors. Prior to conducting regression analysis with the SDM, it is necessary to test and specify the specific form of the model, as shown in

Table 4. Test results of selecting models.

Test	Statistic	p Value	Conclusion
LM (error) test	156.788	0.000	SEM were applicable
Robust LM (error) test	96.320	0.000
LM (lag) test	72.890	0.000	SAR were applicable
Robust LM (lag) test	12.422	0.000
Hausman test	−658.37		Fixed effects were used
LR test assumption: ind nested in both	122.55	0.000	Reject
LR test assumption: time nested in both	735.32	0.000	Reject
Wald test assumption: SDM degradation to SAR	28.62	0.000	Reject
Wald test assumption: SDM degradation to SEM	41.53	0.000	Reject
LR test assumption: SAR nested in SDM	28.50	0.000	Reject
LR test assumption: SEM nested in SDM	41.10	0.000	Reject

. First, the LM test results show that the LM (error) test, Robust LM (error) test, LM (lag) test, and Robust LM (lag) test all passed the test at the 1% significance level, indicating that both the Spatial Error Model (SEM) and Spatial Autoregressive Model (SAR) are applicable. Second, the Hausman test statistic is negative (−658.37). In finite samples, negative Hausman statistics can occur when the estimated variance–covariance matrix difference [Var($\beta$_FE) − Var($\beta$_RE)] fails to be positive semi-definite, a recognized phenomenon in panel econometrics that does not invalidate the test [52]. Following established practice, a negative statistic is interpreted as strong evidence against the random effects specification, thus supporting the use of fixed effects. This interpretation is corroborated by the LR test results (Both-Ind, p < 0.01; Both-Time, p < 0.01), which independently confirm the superiority of dual fixed effects. Third, according to the LR test results, the individual and time dual fixed effects are statistically significantly superior to the individual fixed effects and time fixed effects (Both-Ind, p < 0.01; Both-Time, p < 0.01), implying that the selection of dual fixed effects for the SDM is more reasonable. Finally, the p-values of the two statistics in the Wald test are both less than 0.01, rejecting the null hypothesis, which indicates that the SDM cannot be reduced to the SEM or SAR. Further testing of the LR statistics yielded the same result as the Wald test, revalidating the applicability of the SDM. Therefore, this paper selects the SDM with individual and time dual fixed effects for in-depth analysis (Table 4).

4.4.3. Benchmark Regression Results Analysis

Table 5. Benchmark regression results.

Variable	(1) OLS	(2) SEM	(3) SAR	(4) SDM
struc	−0.028 (−1.10)	−0.126 *** (−4.28)	−0.133 *** (−4.78)	−0.114 *** (−3.90)
finance	0.005 ** (2.34)	0.003 (1.14)	0.003 (1.30)	0.004 (1.60)
gov	−0.317 *** (−11.83)	−0.039 (−1.07)	−0.063 * (−1.86)	−0.014 (−0.41)
npro	0.574 *** (10.21)	0.228 *** (3.18)	0.253 *** (3.58)	0.297 *** (4.14)
open	0.080 *** (4.20)	−0.007 (−0.29)	−0.020 (−0.87)	−0.020 (−0.87)
cons	0.354 *** (20.91)
W × struc				−0.104 * (−1.93)
W × finance				0.004 (1.06)
W × gov				−0.244 ** (−3.95)
W × npro				0.198 (1.43)
W × open				−0.073 * (−1.73)
rho			0.352 *** (9.60)	0.316 *** (8.24)
lambda		0.343 *** (8.84)
sigma2_e			0.003 *** (21.84)	0.025 *** (21.87)
R2	0.797	0.635	0.645	0.652
Log-likelihood		1508.943	1515.248	1529.501
N	975	975	975	975

Note: Z-values are in parentheses; *, **, *** indicate significance levels of 10%, 5%, and 1%, respectively.

presents the regression results of various influencing factors on the CCD between TI and IGG in the YRB under the SDM. From the perspective of model test results, the SDM has a relatively large log-likelihood value, indicating that the overall model is reasonably specified and the estimation results are highly credible. The R-square value is 0.152, which is at a relatively high level, suggesting that the selected variables can well explain the changes in CCD—i.e., they cover the main factors affecting the CCD. The rho of the SDM is 0.337 and statistically significant at the 1% level. This indicates that the model not only involves exogenous interaction effects of explanatory variables but also endogenous interaction of the explained variable, further verifying the applicability of the SDM. Moreover, it implies that the coordinated development of the TI and IGG in the YRB exhibits significant positive spatial spillover effects: the diffusion effects of cities with high coordination levels contribute more to the coordinated development of low-level cities, and spatial spillover is an indispensable key factor in promoting the coordinated development of the two systems.

4.4.4. Decomposition Analysis of Spatial Effects

Constrained by factors such as the feedback effects of the spatial lag term, the direct estimated values of the SDM cannot accurately reflect the true impact magnitude among variables [53]. For this reason, the partial differential method is employed to decompose the impact effects of each variable into direct effects, indirect effects (spillover effects), and total effects, so as to more scientifically reveal the underlying impact mechanism of the coordinated development between TI and IGG in the YRB (

Table 6. Decomposition of SDM effects.

Variable	Direct Effect	Indirect Effect	Total Effect
struc	−0.125 *** (−4.15)	−0.190 *** (−2.73)	−0.314 *** (−4.07)
finance	0.004 * (1.85)	0.008 (1.34)	0.012 * (1.74)
gov	−0.036 (−1.06)	−0.343 *** (−3.99)	−0.379 *** (−3.93)
npro	0.328 *** (4.63)	0.400 ** (2.09)	0.728 *** (3.30)
open	−0.028 (−1.24)	−0.105 * (−1.82)	−0.133 ** (−2.05)

Note: Z-values are in parentheses; *, **, *** indicate significance levels of 10%, 5%, and 1%, respectively.

).

The regression results indicate that the direct effect of industrial structure on the local CCD between TI and IGG is significantly negative ($\beta$ = −0.125, p < 0.01), and its negative spatial spillover effect on adjacent cities also passes the significance test at the 1% level ($\beta$ = −0.190, p < 0.01). The total effect exhibits a significant negative inhibitory characteristic ($\beta$ = −0.314, p < 0.01). This finding reflects two underlying mechanisms. First, the “path dependence and lock-in” mechanism: the YRB’s industrial structure exhibits strong dependence on energy-intensive industries established during earlier industrialization, creating institutional rigidity that constrains green transformation and crowds out resources for tourism development and ecological protection. Second, the “homogeneous competition and externality transmission” mechanism: adjacent cities in the YRB share similar resource endowments (particularly in Shanxi-Shaanxi-Inner Mongolia), leading to convergent industrial structures that intensify inter-city competition and facilitate cross-boundary transmission of environmental pollution, thereby undermining adjacent cities’ tourism attractiveness and ecological quality. This is consistent with the conclusions of existing studies that industrial structure exerts a negative spatial impact on tourism ecological efficiency [54].

The level of financial development exerts a positively significant direct effect on the local CCD at the 10% significance level ($\beta$ = 0.004, p < 0.10), whereas its positive spatial spillover impact on neighboring cities does not pass the significance test ($\beta$ = 0.008, p > 0.10). Meanwhile, the total effect of financial development on CCD is positively significant at the marginal significance level ($\beta$ = 0.012, p < 0.10). This implies that the accumulation and efficient allocation of financial resources can provide capital support for local tourism industry’s green investment and inclusive employment projects, promoting the in-depth integration of the tourism industry and green growth. However, factors such as cross-regional flow barriers of financial resources and regional bias in credit allocation have prevented the full release of its spatial spillover effect. No effective financial synergy support mechanism has been formed among adjacent cities, and this characteristic is consistent with the general laws of spatial spillover in regional financial development [55].

The direct effect of government intervention degree on the local CCD is not statistically significant ($\beta$ = −0.036, p > 0.10), but its negative spatial spillover effect on adjacent cities is significant at the 1% level ($\beta$ = −0.343, p < 0.01), with the total effect showing a significant negative characteristic ($\beta$ = −0.379, p < 0.01). This reflects that local governments’ intervention behaviors may have problems of “local protectionism” and fragmentation in policy implementation. On the one hand, local governments’ regulatory measures have not yet formed an effective synergy in balancing TI development and green growth goals. On the other hand, there is a lack of a collaborative governance mechanism among regions. The differentiated policies adopted by local governments to compete for tourism resources and tax benefits may trigger distortion of factor mobility and conflicts in development paths, thereby exerting an inhibitory effect on the coordinated development of adjacent cities. This is consistent with the intrinsic logical consistency of research conclusions on the spatial governance of tourism ecological efficiency in the YRB [54].

The direct effect of new-quality productive forces on the local CCD is significantly positive at the 1% level ($\beta$ = 0.328, p < 0.01), and its positive spatial spillover effect on adjacent cities passes the significance test at the 5% level ($\beta$ = 0.400, p < 0.05). The total effect exhibits a strong positive driving role ($\beta$ = 0.728, p < 0.01). As a new type of production factor centered on technological innovation and green low-carbon development, new-quality productive forces can not only promote the digital and intelligent transformation of the TI through technological empowerment, improving resource utilization efficiency and the quality of inclusive employment, but also radiate adjacent regions through mechanisms such as technology diffusion and industrial linkage. For example, the cross-regional promotion of smart tourism technologies and shared application of green energy technologies can help form a spatial network of coordinated development within the basin. This corroborates the research conclusions on the positive spatial spillover effect of new-quality productive forces on tourism industry development [21].

The direct effect of opening-up degree on the local CCD is not statistically significant ($\beta$ = −0.028, p > 0.10), while its negative spatial spillover effect on adjacent cities is significant at the 10% level ($\beta$ = −0.105, p < 0.10). The total effect is significantly negative at the 1% level ($\beta$ = −0.133, p < 0.01). This finding can be explained through two theoretical mechanisms. First, the “race-to-the-bottom” mechanism: when regions compete for foreign investment, they may strategically lower environmental standards to enhance locational attractiveness, particularly in upstream and midstream cities facing greater development pressure, thereby undermining the ecological foundation for coordinated development in adjacent areas. Second, the “pollution haven” effect: opening-up in the YRB often involves accepting pollution-intensive industries relocated from more developed regions, creating negative environmental externalities that spill over to neighboring cities through atmospheric and water pollution transmission. Furthermore, the current structure of opening-up in the YRB remains oriented toward traditional resource-based industries rather than green service trade, explaining why opening-up benefits tend to concentrate locally while environmental costs diffuse spatially. This also reveals the characteristics of heterogeneous spatial effects of opening-up in regional coordinated development.

4.4.5. Robustness Tests

To test the robustness of the spatial effect coefficients, this paper adopts two methods: replacing the spatial weight matrix and excluding special years [56]. First, the inverse distance squared spatial weight matrix is used for regression analysis. Second, considering the significant impact of the COVID-19 pandemic on urban economic development and social governance in 2020, data from 2020 to 2022 are excluded to avoid the interference of abnormal years on the research results, and the SDM estimation is conducted again. Subsequently, the decomposition of spatial effects is performed for the regression coefficients of both methods, with the results presented in

Table 7. Results of robustness test.

Variable	Replacing the Spatial Weight Matrix			Excluding Special Years
	Direct Effect	Indirect Effect	Total Effect	Direct Effect	Indirect Effect	Total Effect
struc	−0.113 ***	−0.049	−0.162	−0.139 ***	−0.482 ***	−0.622 ***
	(−3.73)	(−0.49)	(−1.51)	(−4.21)	(−6.05)	(−6.87)
finance	0.006 **	0.010	0.015	0.008 ***	−0.001	0.008
	(2.54)	(0.94)	(1.37)	(3.99)	(−0.01)	(1.06)
gov	−0.064 *	−0.402 ***	−0.477 ***	−0.033	−0.441 ***	−0.474 ***
	(−1.82)	(−3.45)	(−3.69)	(−0.62)	(−3.14)	(−2.89)
npro	0.360 ***	1.077 **	1.437 ***	0.680 ***	−0.786 **	−0.106
	(4.97)	(3.35)	(4.11)	(4.24)	(−1.97)	(−0.23)
open	−0.030	0.020	−0.011	−0.138 ***	−0.294 ***	−0.433 ***
	(−1.32)	(0.22)	(−0.11)	(−4.18)	(−3.20)	(−4.25)
rho	0.355 ***			1.186 ***
	(6.54)			(5.42)
sigma2_e	0.003 ***			0.003 ***
	(21.88)			(21.74)
R2	0.660			0.659
Log-likelihood	1516.031			1498.145

Note: Z-values are in parentheses; *, **, *** indicate significance levels of 10%, 5%, and 1%, respectively.

. It can be observed that among the core explanatory variables, the negative effect of industrial structure, the positive driving effect of new-quality productive forces, the negative spatial spillover effect of government intervention, and the positive local supporting effect of financial development level are consistent with the benchmark regression in terms of effect direction and significance. Only reasonable differences exist in the intensity of some marginal effects and effects related to short-term fluctuations. This finding suggests that the outcomes of spatial econometric analyses regarding the determinants of the CCD for the synergistic coupling coordination between TI and IGG within the YRB exhibit strong robustness. They are not affected by the setting of spatial weights or short-term external shocks, confirming that the previous conclusions are robust and persuasive.

4.4.6. Endogeneity Tests

Considering that the static SDM embodies a spatial weight matrix alongside the spatial lag component of the explained variable, an endogeneity issue may emerge due to interactive effects and reverse causal relationships between the explained variable and explanatory variables. To mitigate such endogeneity-stemming from both the spatial lag component of the explained variable and omitted variables-this study employs a dynamic SDM featuring dual fixed effects (individual and time), a specification that integrates both the temporal and spatial lag components of the explained variable [57]. The findings are reported in

Table 8. Results of endogeneity test.

Variable	Short-Term Effect			Long-Term Effect
	Direct Effect	Indirect Effect	Total Effect	Direct Effect	Indirect Effect	Total Effect
struc	−0.064 **	0.145 **	0.082	−0.065	0.406 *	0.341
	(−2.29)	(2.51)	(1.38)	(−1.52)	(1.68)	(1.30)
finance	−0.002	−0.009	−0.009	−0.005	−0.039	−0.044
	(−0.74)	(−1.37)	(−1.39)	(−1.07)	(−1.34)	(−1.34)
gov	0.019 *	−0.280 ***	−0.261 **	−0.037	−1.059 **	1.096 **
	(0.52)	(−3.12)	(−2.59)	(−0.57)	(−2.21)	(−2.10)
npro	0.246 ***	0.373 *	0.619 ***	0.474 ***	2.122 *	2.595 **
	(3.22)	(1.75)	(2.60)	(3.33)	(1.85)	(2.08)
open	0.018	−0.034	−0.016	0.021	−0.092 ***	−0.071
	(0.72)	(−0.50)	(−0.20)	(0.45)	(−0.28)	(−0.20)
L. CCD	0.257 ***
	(7.39)
L. W × CCD	0.191 ***
	(3.68)
rho	0.402 ***
	(10.98)
sigma2_e	0.003 ***
	(22.67)
R2	0.675
Log-likelihood	1361.180

Note: Z-values are in parentheses; *, **, *** indicate significance levels of 10%, 5%, and 1%, respectively.

. Across both direct and indirect effects, the absolute value of the long-term impact of each explanatory variable exceeds that of its short-term impact. This suggests that every explanatory variable is capable of exerting a far-reaching long-term influence on the CCD, demonstrating a cumulative effect. In the short run, the direct impact of industrial structure is notably negative, while its indirect impact is markedly positive, and the overall effect lacks statistical significance. Though there are differences from the benchmark regression, combined with the short-term adjustment characteristics of the dynamic model, these differences stem from the short-term complementary effect of industrial structure’s spatial transmission after endogeneity control. The long-term effect will still revert to its inherent inhibitory nature, and the core inhibitory logic remains unchanged. In the long term, although some long-term effects of financial development level and industrial structure fail to pass the significance test, there is no essential difference in the direction of effects compared with the benchmark regression, reflecting that their mechanisms are affected by temporal adjustments. Meanwhile, both the first-order lag term of the CCD between TI and IGG in the YRB and its spatial lag term are statistically significant at the 1% level. This indicates that the static SDM has a certain bias due to ignoring unobservable factors. However, the direction of effects of all core explanatory variables is highly consistent with the core conclusions of the benchmark regression, with only reasonable differences in effect intensity and partial marginal significance. This verifies the reliability of the benchmark results and provides a solid empirical foundation for subsequent research conclusions and policy implications.

4.4.7. Heterogeneity Tests

Given that the impact of various factors on the coordinated development between TI and IGG in the YRB may vary across regions, this study further conducts regional heterogeneity tests targeting the upstream, midstream, and downstream regions, with the results presented in

Table 9. Heterogeneity analysis results.

Variable	Upstream		Midstream		Downstream
	Direct Effect	Indirect Effect	Direct Effect	Indirect Effect	Direct Effect	Indirect Effect
struc	−0.163 ***	0.228 ***	−0.052	−0.338 ***	−0.126 **	−0.428 **
	(−2.88)	(2.96)	(−1.12)	(−3.92)	(−2.28)	(−2.43)
finance	0.011 ***	0.001	0.002	−0.013	−0.004	0.023 *
	(3.10)	(0.18)	(0.59)	(−1.37)	(−1.02)	(1.81)
gov	−0.019	−0.223 **	−0.059	0.645 ***	0.124 *	−0.605 **
	(−0.35)	(−2.55)	(−0.67)	(2.91)	(1.68)	(−2.34)
npro	−0.146	0.787 **	0.636 ***	2.697 ***	0.568 ***	0.056
	(−0.78)	(2.42)	(4.19)	(6.65)	(6.14)	(0.21)
open	0.002	−0.107	−0.085 ***	−0.260 ***	0.053	−0.026
	(0.03)	(−0.69)	(−2.59)	(−3.67)	(1.46)	(−0.26)

Note: Z-values are in parentheses; *, **, *** indicate significance levels of 10%, 5%, and 1%, respectively.

. Specifically, there is significant regional heterogeneity in the effects of each influencing factor on coordinated development, showing a differentiated correspondence with the overall basin effect. In the upstream region, the industrial structure exhibits a unique characteristic of “local inhibition and adjacent promotion” (direct effect: $\beta$ = −0.163, p < 0.01; spillover effect: $\beta$ = 0.228, p < 0.01), which contrasts with the dual negative effects of the entire basin. This may stem from the complementary division of labor between the energy industry in the upstream and eco-tourism in adjacent cities. Although the local effect of new-quality productive forces is not statistically significant ($\beta$ = −0.146, p > 0.10), its positive spillover effect is significant ($\beta$ = 0.787, p < 0.05), reflecting the phased characteristic that technology diffusion precedes local absorption and digestion. In the midstream region, both the direct effect ($\beta$ = 0.636, p < 0.01) and spillover effect ($\beta$ = 2.697, p < 0.01) of new-quality productive forces far exceed the overall basin level, making them the core engine of regional coordinated development. In contrast, the opening-up presents the characteristic of “dual inhibition in local and adjacent regions”, highlighting the path lock-in of traditional export-oriented industries in the midstream. The factor competition and transmission of environmental externalities exert dual inhibition on the green coordinated development of local and surrounding cities. In the downstream region, the dual negative effects of industrial structure ($\beta$ = −0.126, p < 0.05; $\beta$ = −0.428, p < 0.01) contrast with the “local promotion and adjacent inhibition” of government intervention ($\beta$ = 0.124, p < 0.10; $\beta$ = −0.605, p < 0.05). This reflects the contradiction between industrial upgrading pressure and policy space competition in developed regions. The root cause of regional heterogeneity lies in differences in development foundations: the industrial complementarity of upstream ecological function zones, the technological agglomeration of midstream industrial bases, and the resource competition of downstream economic core zones lead to the differentiation of the action paths of the same factor. This verifies the theoretical hypothesis of the “factor endowment-determined mechanism” in regional economic geography and provides empirical evidence for differentiated governance of the river basin [58].

5. Discussion

5.1. Core Findings and Theoretical Implications

First, it verifies the applicability of the coupling coordination theory at the river basin scale. The coordinated development level of the YRB shows a fluctuating upward trend, presenting a “downstream-midstream-upstream” gradient-decreasing pattern and a dual-core radiation pattern in spatial distribution. This reflects the interaction law between the TI system and IGG system. This finding expands the application of the coupling coordination theory from the national and provincial scales to the river basin scale, enriching the theoretical understanding of the spatial differentiation of coordinated development under the constraints of ecological security and economic development.

Second, it deepens the understanding of the spatial spillover mechanism in the context of river basin development. Based on the new economic geography theory, this paper identifies that there is significant positive spatial autocorrelation and agglomeration characteristics in coordinated development and clarifies the heterogeneous spatial spillover effects of various influencing factors—new-quality productive forces have significant positive direct effects and spillover effects, while industrial structure has significant negative direct effects and spillover effects. This conclusion supplements the empirical evidence of spatial spillover effects in river basin regions, reveals the dynamic game between siphon effects and diffusion effects in the process of coordinated development, and improves the theoretical system of spatial economics.

Third, it expands the research perspective on the relationship between TI and IGG. Existing studies mostly focus on the national or provincial scales. Taking the YRB, a compound region of “ecological security barrier-economic development belt”, as the research object, this study explores the interaction mechanism from the perspective of spatial correlation. It is found that TI can promote IGG through scale-driven, structure-optimizing, and potential-leading effects, while IGG can feed back into the development of the TI through economic support, social protection, people’s livelihood driving, and ecological regulation effects. This finding deepens the theoretical understanding of the mutual empowerment relationship between tourism and sustainable development.

The underlying reasons for these spatial patterns can be attributed to three interrelated mechanisms. First, the “downstream–midstream–upstream” gradient reflects the historical accumulation of economic capital, human resources, and institutional capacity. Downstream regions, particularly the Shandong Peninsula, have benefited from early opening-up policies and coastal location advantages, establishing mature tourism industries and robust green development foundations. Second, the dual-core radiation pattern emerges from agglomeration economies—core cities attract high-quality factors through superior infrastructure, policy efficiency, and market access, subsequently diffusing development benefits to surrounding areas through demonstration effects and industrial linkages. Third, the persistent LL clusters in upstream regions result from the compounding effects of ecological sensitivity constraints, underdeveloped infrastructure, and limited absorptive capacity for technological spillovers. These mechanisms collectively explain why coordinated development in the YRB exhibits such pronounced spatial heterogeneity and why differentiated policy interventions are essential.

5.2. Comparison with Previous Research

This study’s findings align with prior scholarly work and further offer supplementary insights and extensions to existing research. On the one hand, the negative impact of industrial structure on coordinated development is consistent with the research of Tian et al. [16], who pointed out that an irrational industrial structure restricts the green development of the TI. However, this paper further identifies that industrial structure has a significant negative spatial spillover effect, which stems from homogeneous competition in industrial layout among adjacent cities, expanding the cognitive boundary of the industrial structure’s impact mechanism. On the other hand, the positive effect of new-quality productive forces on coordinated development is logically consistent with the research on digital infrastructure by Liu et al. [21]. Nevertheless, this paper clarifies the dual positive effects of new-quality productive forces on local and adjacent regions, highlighting the importance of technology diffusion in river basin coordinated development.

Regarding YRB-specific research, this study both corroborates and extends existing findings. First, the “downstream–midstream–upstream” gradient pattern is consistent with Cai et al. [30], who identified similar spatial differentiation in tourism-ecology coordination within the YRB. However, this study further reveals the dual-core radiation mechanism centered on the Jinan–Qingdao and Xi’an–Zhengzhou agglomerations, providing new insights into the spatial organization of coordinated development. Second, the significant positive spillover effect of new-quality productive forces ($\beta$ = 0.400, p < 0.05) aligns with Zhang et al. [31], who emphasized the role of ICT in promoting tourism economic development in the YRB. Nevertheless, this study advances their work by identifying the “spillover-preceding-absorption” pattern in upstream regions, suggesting that technology diffusion precedes local absorption capacity development—a finding with important policy implications for technology transfer strategies. Third, while Tian et al. [54] confirmed the negative impact of industrial structure on tourism ecological efficiency in the YRB, this study further identifies its significant negative spatial spillover effect ($\beta$ = −0.190, p < 0.01), revealing the cross-regional transmission mechanism through homogeneous competition and environmental externalities. These comparisons demonstrate that the TI-IGG coordinated development in the YRB exhibits unique spatial characteristics requiring targeted interventions distinct from national-level policies.

The differences between this study and previous YRB research can be attributed to three factors. First, methodological advancement—while previous studies such as Cai et al. [30] employed traditional coupling coordination analysis without spatial decomposition, this study adopts the SDM to separately identify direct and indirect effects, enabling the detection of spatial spillover mechanisms that were previously unobservable. Second, variable innovation—the inclusion of new-quality productive forces as a key explanatory variable, which was not examined in previous YRB tourism studies, reveals technology diffusion patterns unique to the basin context. Third, temporal extension—this study covers the period up to 2023, capturing recent policy effects from the Yellow River Basin ecological protection and high-quality development strategy implemented since 2019, which earlier studies could not incorporate.

In addition, the conclusion of this paper that core influencing factors exhibit significant regional heterogeneity aligns with the view of Zhao et al. [58], who emphasized the impact of factor endowments on regional development differences. However, this paper further identifies the differentiated effects of various factors in the upstream, midstream, and downstream regions, providing more detailed empirical evidence for differentiated governance. Compared with existing studies that ignore spatial effects, this study adopts the SDM to decompose direct effects and indirect effects, avoiding estimation bias caused by the spatial homogeneity assumption and enhancing the reliability of the conclusions.

5.3. Practical Implications and Alignment with SDGs

This study’s findings carry notable practical implications for advancing the implementation of the national strategy for ecological conservation and high-quality development in the YRB, and for the attainment of the SDGs. In terms of alignment with SDGs, the paper’s focus on the coordinated development between the TI and IGG directly responds to SDG 8 (Decent Work and Economic Growth), SDG 13 (Climate Action), SDG 11 (Sustainable Cities and Communities), and SDG 1 (No Poverty). The proposed strategies, such as promoting inclusive tourism development, strengthening ecological protection, and optimizing the industrial structure, provide practical pathways for the regional implementation of SDGs.

For the YRB, the conclusions offer a scientific basis for addressing the fragmentation of river basin governance and achieving integrated development: the dual-core radiation pattern of the Jinan-Qingdao and Xi’an-Zhengzhou urban agglomerations points out the direction for driving the coordinated development of the entire basin relying on core cities. The negative spillover effects of government intervention and opening-up highlight the need to strengthen cross-regional collaboration and avoid local protectionism and vicious competition. For similar river basins and ecologically sensitive regions worldwide, the research framework and policy recommendations of this study have reference value. In particular, the practical experience in balancing ecological protection and economic development is worthy of global promotion.

6. Conclusions

6.1. Main Conclusions

Addressing the needs of implementing the national strategy for ecological protection and high-quality development of the YRB, this paper explores the spatiotemporal evolution characteristics and spatial spillover effects of the coordinated development between TI and IGG in the YRB from 2011 to 2023. The main conclusions are as follows:

(1)

In terms of temporal evolution: The level of coordinated development between TI and IGG in the YRB fluctuated and improved from 2011 to 2023, with a corresponding improvement in the coordinated development level.

(2)

In terms of spatial distribution: The coordinated development between TI and IGG in the YRB is in an unbalanced state, presenting an overall gradient-decreasing pattern of “downstream-midstream-upstream”. Cities with high coordinated development levels have gradually formed dual-core radiation belts centered on the Jinan-Qingdao and Xi’an–Zhengzhou urban agglomerations, but their radiation capacity to peripheral cities remains limited.

(3)

In terms of spatial correlation: There is significant positive spatial autocorrelation in the coordinated development between TI and IGG in the YRB. The spatial agglomeration characteristics have weakened in a fluctuating manner over time, and a convergence trend has emerged in coordinated development. However, the polarization phenomenon remains prominent, forming a typical Matthew effect of “the strong becoming stronger and the weak becoming weaker”. HH clusters are mainly concentrated in the Shandong Peninsula in the downstream, while LL clusters are primarily distributed in the Qinghai-Gansu-Ningxia-Inner Mongolia region in the upstream.

(4)

The SDM regression results indicate that: Industrial structure exerts significant inhibitory effects on the coordinated development between TI and IGG both in local and adjacent cities, serving as a key constraint limiting the improvement of coordinated development levels. Financial development level significantly promotes local coordinated development, but its spatial spillover effect is not statistically significant. New-quality productive forces not only improve the coordinated development level of local cities but also enhance that of adjacent cities, generating significant positive spatial spillover effects. Government intervention degree and opening-up level have no significant impacts on local coordinated development but exert significant inhibitory effects on the coordinated development of adjacent cities, with prominent negative spatial spillover effects. The heterogeneity test results show that the core factors affecting the basin’s coordinated development exhibit significant regional heterogeneity, which is closely related to differences in factor endowments and differs from the overall effect.

6.2. Policy Implications

To promote the coordinated development of the TI and IGG in the YRB, targeted policy implications are proposed as follows:

These recommendations align with national strategic frameworks. The Outline of the Plan for Ecological Protection and High-Quality Development of the Yellow River Basin (2021) emphasizes “adjusting regional industrial layout and confining economic activities within the carrying capacity of resources and environment” and advocates differentiated tourism development for the upstream, midstream, and downstream regions [34]. The “14th Five-Year Plan” for Tourism Development (2022) proposes to “improve the tourism coordination mechanism for the Yellow River Basin” and promote technology application in tourism [59].

First, construct a “Dual-Core Leadership-Gradient Linkage” spatial governance system by strengthening the radiation capacity of the Jinan-Qingdao and Xi’an-Zhengzhou urban agglomerations—downstream core cities should enhance the integration quality of the TI and green industries through cross-city platforms for technology transfer, talent sharing, and resource allocation via the “point-axis” diffusion model, while midstream core cities need to integrate “Silk Road starting point” cultural-tourism resources with low-carbon policies to build a coordination hub connecting upstream and downstream and mitigate the “core-periphery” imbalance. Remove administrative barriers to advance cross-regional collaborative governance, establish a unified management agency for the YRB’s coordinated development, and formulate consistent development plans, ecological compensation standards, and inter-regional industrial cooperation mechanisms; implement differentiated policies for upstream LL clusters by increasing investment in tourism infrastructure and ecological protection, and offering tax reductions and subsidies to attract green tourism investment, thereby avoiding “low-level lock-in” path dependence and promoting leapfrog upgrading of coordination levels.

Furthermore, accelerate the green transformation of industrial structure to address negative spatial spillovers. Our finding of significant negative spillover effects ($\beta$ = −0.190, p < 0.01) empirically validates the Outline’s requirement to “adjust regional industrial layout” [34]. Specific measures should include establishing region-differentiated environmental access standards with stricter thresholds for ecologically sensitive upstream areas; creating cross-regional coordination mechanisms to reduce homogeneous competition that intensifies negative spillovers, consistent with the Outline’s call for “coordinated ecological protection” [34]; and promoting the transition to high-value-added tourism formats such as cultural-tourism integration and smart tourism, following the Outline’s differentiated regional positioning for upstream, midstream, and downstream areas [34].

Additionally, cultivate new-quality productive forces to strengthen positive spillovers and enable upstream regions to benefit from technology diffusion. Our heterogeneity analysis reveals a notable “spillover-preceding-absorption” pattern in the upstream: strong spillover effects ($\beta$ = 0.787, p < 0.05) despite non-significant local effects ($\beta$ = −0.146, p > 0.10), suggesting that technology diffusion precedes local absorption capacity development. To capitalize on this pattern, we propose: establishing technology transfer channels between core cities and upstream regions, aligning with the “14th Five-Year Plan”’s emphasis on “applying new technologies such as big data, cloud computing, and 5G in tourism” [59]; implementing talent exchange programs to enhance absorption capacity; and prioritizing digital infrastructure deployment in upstream scenic areas, as the “14th Five-Year Plan” calls for “enhancing 5G network coverage in key tourism areas” [59]. These measures respond to the Outline’s requirement to “strengthen supporting infrastructure construction” in the upstream region [34].

Finally, rationalize factor allocation with a view to mitigating adverse spatial spillover impacts: enhance the efficiency of spatial allocation of financial resources through the promotion of green tourism credit instruments and cross-regional financial collaboration, so as to dismantle barriers to resource mobility; regulate government intervention to avoid local protectionism, coordinate inter-regional tourism policies, and guide local governments toward indirect regulation (e.g., ecological supervision and public service provision) to reduce policy fragmentation impacts; promote high-level opening-up by prioritizing the introduction of high-end tourism services and green technologies, avoiding low-level redundant construction, and strengthening international eco-tourism and cultural tourism cooperation to enhance the YRB’s tourism competitiveness.

To enhance the operationality of the above recommendations,

Table 10. Structured policy framework aligned with empirical findings.

Time Horizon	Key Empirical Finding	Target Region	Targeted Policy Measure
Short-term (1–2 years)	Government intervention: negative spillover ($\beta$ = −0.343 ***)	Basin-wide	Establish inter-city policy coordination committees; unify environmental standards
	New-quality productive forces: spillover-preceding-absorption pattern (spillover $\beta$ = 0.787 **; local $\beta$ = −0.146)	Upstream	Priority 5G and digital infrastructure deployment; technology demonstration centers
Medium-term (3–5 years)	Industrial structure: negative spillover ($\beta$ = −0.190 ***)	Midstream	Industrial differentiation policies; cross-regional ecological compensation fund
	Opening-up: negative spillover ($\beta$ = −0.105 *)	Basin-wide	Shift FDI attraction toward green service industries
Long-term (5–10 years)	New-quality productive forces: positive total effect ($\beta$ = 0.728 ***)	Basin-wide	Integrated smart tourism network connecting dual cores with upstream zones
	LL cluster persistence	Upstream	Systematic infrastructure upgrading and talent cultivation programs

Note: *, **, *** indicate significance levels of 10%, 5%, and 1%, respectively.

presents a structured policy framework that aligns specific actions with our empirical findings across different time horizons. Short-term actions (1–2 years) should address immediate coordination failures: given the significant negative spillover of government intervention ($\beta$ = −0.343, p < 0.01), establishing inter-city policy coordination mechanisms is essential; for upstream regions showing strong technology spillovers but weak local absorption (spillover β = 0.787, p < 0.05; local $\beta$ = −0.146, p > 0.10), priority digital infrastructure deployment can enhance absorptive capacity. Medium-term actions (3–5 years) should focus on industrial restructuring to address negative spillovers ($\beta$ = −0.190, p < 0.01) through differentiated regional positioning and ecological compensation mechanisms. Long-term actions (5–10 years) should leverage new-quality productive forces’ strong positive effects (total $\beta$ = 0.728, p < 0.01) to build an integrated smart tourism network across the basin.

6.3. Limitations and Future Directions

While this study makes theoretical and practical contributions, it inevitably has certain limitations: First, regarding indicator selection, the tourism subsystem primarily depends on objective metrics—including tourism revenue and tourist reception volume—and lacks subjective measures such as tourist satisfaction and residents’ sense of acquisition, a gap that might undermine the comprehensiveness of the assessment results. Second, this study focuses on the direct effects of influencing factors and fails to explore the mediating and moderating mechanisms of variables such as digital economy and environmental regulation. Third, the research period is 2011–2023, and the long-term dynamic evolution mechanism requires support from longer time-series data.

Future research can be carried out from five aspects: First, improve the evaluation index system, integrate subjective and objective dimensions, and combine mixed research methods such as questionnaires and interviews to achieve a comprehensive and accurate evaluation of coordinated development. Second, further explore the mediating and moderating mechanisms of key variables such as digital economy and social capital to deepen the understanding of the impact paths of coordinated development. Third, expand the research spatial scope, compare the coordinated development characteristics of typical river basins such as the Yangtze River Basin and the Rhine River Basin, and reveal the universal laws of river basin sustainable development. Fourth, adopt dynamic spatial econometric models to explore the long-term evolutionary trends and driving factors of coordinated development so as to provide more targeted policy implications. Fifth, conduct sensitivity analysis using alternative weighting methods (such as equal weighting, PCA-based weights, and expert-assigned weights) to further verify the robustness of the evaluation results, thereby enhancing the methodological rigor of coupling coordination research.