Revisiting Tourism Development and Economic Growth: A Framework for Configurational Analysis in Chinese Cities

Abstract

This paper comparatively analyzes the sufficiency and necessity of tourism’s influence on economic growth in different cities from a systematic configurational perspective. Two important time points in China’s tourism development, 2010 and 2019, are also considered in this paper to explore whether the impact of tourism on urban economic growth is temporally heterogeneous. The results demonstrate that tourism is not necessary for urban economic growth. However, the dependence on the tourism economy plays an important role in several urban economic growth patterns. Only one tourism-driven economic growth pattern exists, where tourism drives economic growth led by investment, and this pattern did not change significantly from 2010 to 2019. A tourism-driven low economic growth model also suggests that a high dependence on tourism leads to low economic growth. Two tourism-constrained low economic growth patterns exist: investment–industrial structure tourism-constrained and investment–innovation tourism-constrained. These two patterns indicate that economic growth rates are difficult to increase if the tourism economy is underdeveloped. In addition, tourism-driven or -constrained economic growth patterns have specific spatial clustering characteristics. This paper argues that tourism should actively seek foreign capital utilization and fixed asset investment, and also constantly reduce its independence and blur its industrial boundaries to better integrate or link with other industries to play its economic growth role. Furthermore, city policymakers should be fully aware of their own (tourism) resource endowment and the internal and external environment changes to choose a suitable economic growth model.

1. Introduction

Urban tourism is a fundamental component of the global tourism industry and holds a crucial position in achieving sustainable development. Consequently, the relationship between urban tourism and economic growth represents an essential aspect of sustainable tourism studies. The positive impact of tourism on regional economic growth has been a prevalent topic in numerous studies [1,2,3,4,5,6]. Such findings provide a solid theoretical basis for promoting tourism at the local level. However, relying solely on statistical significance could lead to the false impression that promoting tourism is a quick fix for economic growth, which might lead to a scenario where the dependence on tourism as the primary source of income results in sluggish growth or a limited level of economic advancement [7]. This scenario is particularly noticeable in China, where urban data indicate a noticeable gap between the level of urban economic expansion and tourism dependence in the region.

For example, the degree of tourism specialization (measured by the proportion of total tourism revenue to GDP) in economically advanced cities such as Guangzhou and Shenzhen were merely 18.85% and 6.37%, respectively, in 2019, whereas prominent destinations such as Huangshan and Guilin exhibited high tourism specialization rates of 80.62% and 89.00%, though their per capita GDPs were only ¥57,853 and ¥41,294. The economic growth rates of 6.54% and 7.74% in Huangshan and Guilin are comparable to the 6.80% in Guangzhou and 6.70% in Shenzhen in 2019. However, certain cities with a specialization in tourism, such as Hezhou and Lijiang (with growth rates of 11.80% and 9.9%, respectively), have notably higher figures (The information in this section was obtained from the 2019 statistical bulletin on national economic and social development for every city.).

Based on the theoretical perception that tourism promotes economic growth, this paper aims to clarify what kind of tourism development can promote economic growth or under what context tourism can promote economic growth. That is, the sufficiency of tourism development for economic growth and its contextual dependence. Is tourism development necessary for economic growth, and how necessary is it? The two questions correspond to two typical variable relationships in social science research: necessity and sufficiency [8]. From an established econometric perspective, the influence of tourism on economic development can be viewed as average effects. While this has significant value in guiding macroeconomic decisions, it may not always fully reflect the nuances of specific cases. Therefore, it is imperative to establish a clear connection between tourism and economic development through a comparative examination of diverse instances, necessitating the exploration of alternative research methodologies distinct from conventional regression analysis.

This paper employs the methodology of fuzzy-set qualitative comparative analysis (fsQCA). This method involves analyzing various cases and examining how the antecedent variables affect the outcome variables and their corresponding solutions (i.e., configuration). This approach takes into account the unique characteristics of each case [9]. The configuration is characterized by a combination of different antecedent variables, yielding a symbiotic, commensal, or competitive relationship between the variables. This approach is additionally appropriate for examining the necessity of antecedent factors for the outcome variable. Therefore, the application of the fsQCA methodology can effectively address the questions of necessity and sufficiency mentioned before. In measurement of necessity, there is a quantitative evaluation, along with the fsQCA approach. In contrast to the fsQCA approach, which is a qualitative test of necessity, the necessity condition analysis (NCA) technique can quantitatively assess the degree of necessity between variables [10]. As a result, it has become increasingly popular for the investigation of necessary relationships. Additionally, the NCA method can gauge the bottleneck level of the antecedent variable, denoting the minimal level the antecedent variable must achieve to produce a specific outcome variable level [11]. For this particular study, the focus is on determining the appropriate level of tourism development that must be met to achieve a specific level of economic growth.

The article introduces two primary innovations. First, this paper diverges from the typical examination of tourism’s average impact on economic growth by instead concentrating on the necessity and sufficiency of tourism in influencing economic growth. Second, this study incorporates the NCA approach with the fsQCA method, effectively addressing the latter’s inadequacy in assessing necessity. This paper makes two theoretical contributions. Firstly, from a configurational perspective, this study extensively and systematically incorporates ecological elements related to tourism and investigates how the coexistence of these elements promotes economic growth in urban areas. It also reveals the various paths through which tourism influences economic growth, as well as the underlying mechanisms and spatial and temporal differences in a symmetrical manner. Secondly, this study investigates whether and to what degree isolated environmental factors related to tourism contribute to urban economic development, thus expanding the current theoretical understanding of the causal relationship between tourism and economic growth.

2. Literature Review

The impact of tourism on regional economic growth has been a prevalent topic in several studies, which have confirmed the positive effects of tourism on economic development [1,2,3,4,5,6]. These studies have also explored the causal relationship between tourism development and economic growth using Granger causality analysis [12,13,14]. However, it has been proved that there is no inevitable positive link between tourism and economic growth. Therefore, the promotion of tourism must involve a holistic approach to achieve sustainable economic growth that is not solely reliant on tourism. As such, the prudent promotion of tourism, coupled with other economic strategies, is necessary to attain sustainable economic development and achieve lasting success in the tourism sector.

According to configuration theory, outcomes result from a combination of factors, and there can be several functionally equivalent paths that lead to high or low outcome variables [15]. The existence of functionally equivalent paths is determined by the manner in which the various elements of the system are integrated. Configuration theory can identify both the shared patterns of outcome generation and the unique aspects of these patterns. Taking into consideration the concept of configuration, the influence of tourism on economic advancement should not merely focus on the tourism industry itself but must also adopt a systemic approach to investigate the synergistic effects of tourism with other significant factors. The diverse impacts of tourism on economic growth are contingent on its interaction with other factors in theory. In summary, configuration theory posits that the complex relationships between economic factors and their fluctuations significantly impact the efficacy of economic systems. However, these connections that involve aspects of necessity, the coupling of multi-condition variables, and asymmetry cannot be adequately examined using standard reductionist research methodologies such as regression analysis.

For instance, the promotion of tourism can improve the destination’s brand reputation and appeal, enhance its level of communication and connectivity, foster the sharing of scientific and technological advancements, and facilitate the attraction of foreign investments [16,17]. Tourism growth can also boost the number of visitors and stimulate related tourism services, diminishing the reliance on conventional agriculture and secondary industries and increasing the portion of tertiary industries, such as services, within the national economy, which, in turn, aids in the enhancement of the industrial structure [18]. Tourism may have a crowding out effect on traditional or other industries due to low barriers to entry for entrepreneurship and employment behavior [19]. Additionally, as tourism is not classified as a science and technology innovation industry [20], it may limit overall innovation capacity in the region and constrain further economic growth. However, the theoretical mechanism underlying the relationship between tourism and market factors, including industrial structure and technological innovation and their systematic impact on regional economic growth, remains unclear. The methodology for configuring tourism environments presents a range of theoretical configurations for different ecologies, providing insight into which configurations may lead to high or low economic growth [21]. Therefore, configuration theory can efficiently support the examination of the intricate sufficiency of tourism’s influence on economic development and its context-specific nature.

In summary, this paper investigates how tourism development can drive economic growth, as analyzed through the framework of configuration theory and employing an integrated approach of fsQCA and NCA methodologies. This paper focuses on two fundamental theoretical inquiries about the necessity and sufficiency of tourism in promoting economic development. The necessity issue concerns whether tourism development is necessary for regional economic growth and how to assess this necessity. The sufficiency query arises within the context of whether tourism can facilitate or hinder economic development—known as the tourism approach for economic growth. Since provincial-level data may overlook intra-regional variations and still be susceptible to the flaws of averaging effects, and county-level data can be hard to acquire, the database used in this paper draws from highly detailed Chinese prefecture-level city data.

3. Methodology

The fsQCA approach is the most popular method among various QCA techniques [22]. The fsQCA method tackles two primary inquiries. One is the necessity of tourism and other antecedent variables for economic growth, and the other is what combination of tourism and other variables can lead to high or low economic growth. This paper employs fsQCA 3.0 software to apply the fsQCA method. The NCA methodology is developed to evaluate the necessity of antecedent factors. It aims to examine hindrances, limitations, and obstacles that impede the realization of desired results. The NCA package, developed by Dul et al. in the R 4.2.1 software [11,13,14], serves as the analysis tool for NCA. According to the results of necessity analysis using fsQCA, the NCA approach enables the validation and further investigation of necessary conditions.

In utilizing the framework of configuration theory, the initial step is to precisely define the boundaries of the system under examination. The topic of this academic research paper identifies tourism as the primary antecedent variable. Furthermore, numerous scholars have pinpointed several crucial factors that greatly impact economic development. The antecedent variables being considered are foreign investment [23,24], innovation [25,26], human resources [27,28], industrial structure [29,30], trade openness [31,32], and fixed asset investment [33,34]. It is vital to note that as economic growth is a complex system, there exist various other factors that significantly impact it, including regional circumstances and the environment. It is pertinent to note that variables are always subject to change, and as such, alterations in local circumstances and the natural environment may not always be noteworthy. Another significant factor is the need to improve decision-making processes, especially as adapting to alterations in the local context and natural environment can prove challenging at decision-making levels. One downside of this research is the limited number of variables available for configuration analysis. An excessive number of variables can result in an overwhelming complexity of possible solutions, whereby the total number of theoretical solutions surpasses the number of actual cases [22]. Ultimately, this paper rigorously identifies seven antecedent variables of economic growth, including tourism, foreign investment, innovation, human resources, industrial structure, trade openness, and fixed asset investment. In line with the configuration theory framework, various complex, symbiotic, and commensal relationships exist among these variables, all of which significantly impact the economic growth performance. In particular, we focus on the potential impact of the complex association between tourism and other antecedent variables on economic growth. The theoretical analysis framework of this paper is illustrated in Figure 1.

Economic growth is commonly measured in terms of GDP per capita [9] and GDP growth rate [35]. Actually, per capita GDP is a better indicator of the level of economic development and does not capture the change over time, whereas GDP growth rate is a better indicator of the impact of policy making on economic development. Thus, this paper employs the GDP growth rate as a measure of economic growth. This paper utilizes the term “tourism specialization” to denote the degree of tourism advancement or development, characterized by the proportion of tourism earnings to GDP [35,36]. This ratio is commonly used to measure tourism specialization in China, as statistical data on tourism value added is limited. Notably, tourism revenue in this study encompasses both inbound and domestic receipts. The paper utilizes a variety of factors to gauge human resources, such as population [37], higher education, and employment [38], represented by urban population density, number of higher education institutions per 100,000 individuals, and year-end employment, respectively. It is obvious that all the three indicators positively affect the level of human resources, so this paper constructs the interaction term of the above three factors to represent the level of human resources.

Innovation is measured by two metrics, namely innovation input and output [38]. Innovation input is gauged by the proportion of scientific input to GDP, while the innovation output is evaluated by the quantity of granted patents [38]. Likewise, innovation input and output have a favorable impact on the degree of innovation. This study creates the interaction term of innovation input and output to depict regional innovation. Foreign trade is typically assessed based on the ratio of imports and exports to GDP [39], while foreign investment is measured by the amount of foreign investment actually used, as indicated by the ratio of such investment to GDP [40]. The primary method for economic development lies in enhancing the share of tertiary industry. As a result, the industrial structure is shaped by the proportion of value added by tertiary industry as compared to that added by secondary industry [41]. Fixed asset investment is measured as the ratio of the amount of fixed assets invested to GDP.

It is important to note that the classical fsQCA approach, despite its unique strengths in necessity and sufficiency analyses, has a natural deficiency in dynamic analysis [42], while economic growth patterns tend to be dynamic. In order to better capture the dynamic changes in the impact of tourism on economic growth, this paper uses a time series fsQCA and NCA approach, i.e., based on homogeneous cases but observing changes in necessity and sufficiency at different points in time. Two time points are identified in this paper: 2010 and 2019. On 25 November 2009, China’s State Council issued the Opinions on Accelerating the Development of Tourism, stating that tourism should be cultivated into a strategic pillar industry of the national economy and a modern service industry that the people are more satisfied with. So, 2010 is an important time for the development of tourism in China. In March 2018, the former Chinese Ministry of Culture and the National Tourism Administration merged into the Ministry of Culture and Tourism, and the administrative mechanism of reform had a huge impact on the development of tourism, thus taking 2019 as the second time point. The examination of landmark tourism development milestones helps to better investigate the impact of tourism and its related variables on economic growth.

Data on tourism revenue for prefecture-level cities are obtained from the 2010 and 2019 national economic and social development statistical bulletins for each city. Data for all other variables are obtained from the 2011 and 2020 China City Statistical Yearbook [43,44]. A total of 272 prefecture-level cities’ variable data are obtained in this paper, and the descriptive statistical results of each variable are shown in

Table 1. Results of descriptive statistics.

Variable	Unit	Mean		STD. DEV		Min.		Max.
		2010	2019	2010	2019	2010	2019	2010	2019
Economic growth	%	14.58	6.44	2.16	1.98	9.60	−3.60	25.10	11.80
Tourism	/	0.1004	0.3088	0.0896	0.2447	0.0091	0.0331	0.7832	2.2796
Foreign capital	/	0.0182	0.0128	0.0178	0.0146	0.0000	0.0000	0.1277	0.0613
Human resources	/	9821	17,950	403,601	75,395	0	0	404,020	786,890
Innovation	/	6.6447	52.9559	26.3843	232.3673	0.0085	0.0058	348.8300	3393.3000
Industrial structure	/	0.7502	1.3722	0.3640	0.6558	0.0833	0.4270	3.0605	4.9456
Fixed asset investment	/	0.7263	0.8710	0.2530	0.3535	0.1063	0.1882	1.6451	2.4108
Trade openness	/	0.2190	0.1632	0.5633	0.2616	0.0002	0.0007	7.8462	2.4834

.

Prior to conducting fsQCA analysis, it is essential to calibrate the variable data. The key objective of calibration is to identify the anchor point for each variable. As the variables in this study, such as outcome and antecedent variables, lack theoretical high or low judgment criteria, the article looks to comparable research and utilizes the data percentile as the calibration reference point [38,45]. This paper uses the 95th percentile as the anchor point for fully in, the median as the crossover point, and the 5th percentile as the anchor point for fully out. The calibration anchor points for each variable are shown in

Table 2. Calibration anchor points.

Variable	Fully in		Crossover Points		Fully Out
	2010	2019	2010	2019	2010	2019
Economic growth	18.40	9.00	14.20	6.80	11.50	2.80
Tourism	0.1996	0.7285	0.0807	0.2465	0.0229	0.0743
Foreign capital	0.0551	0.0470	0.0129	0.0064	0.0003	0.0001
Human resources	44,076	84,429	697	1142	42	62
Innovation	30.7825	199.5719	0.5437	6.5965	0.0489	0.2660
Industrial structure	1.2287	2.6402	0.685	1.2090	0.3444	0.6855
Fixed asset investment	1.1770	1.4371	0.7008	0.8595	0.3529	0.3096
Trade openness	0.7412	0.5520	0.0766	0.0708	0.0068	0.0070

. The calibration outcomes are employed for NCA analysis to enable the comparison of the findings of the necessity analysis.

4. Results

This paper constructs the following fsQCA model.

Economic growth = f (tourism, foreign investment, human resources, innovation, industrial structure, fixed asset investment, import and export)

~Economic growth = f (tourism, foreign investment, human resources, innovation, industrial structure, fixed asset investment, import and export)

where the symbol “~” represents negation or weakness. This paper aims to investigate the necessary and sufficient conditions for high economic growth, as well as those for low or insufficient economic growth.

4.1. Necessity Analysis

In the fsQCA approach, the necessity analysis considers both positive and negative responses of the outcome variable, as well as the presence or absence of the antecedent variable. If the antecedent variable has a consistency of over 0.9, it is deemed necessary for the outcome variable.

Table 3. Analysis of necessary conditions.

Condition	Economic Growth (Consistency)		~Economic Growth (Consistency)
	2010	2019	2010	2019
Tourism	0.6603	0.6328	0.6325	0.6052
~tourism	0.6671	0.6670	0.6923	0.7159
Foreign capital	0.6285	0.6380	0.5919	0.5395
~foreign capital	0.6841	0.6454	0.7182	0.7640
Human resources	0.6030	0.5716	0.6032	0.6198
~human resources	0.7322	0.7418	0.7293	0.7160
Innovation	0.5247	0.5742	0.6376	0.5715
~innovation	0.8132	0.7304	0.6976	0.7548
Industrial structure	0.6374	0.5913	0.6926	0.6761
~industrial structure	0.7082	0.7311	0.6502	0.6693
Fixed asset investment	0.6909	0.7154	0.6270	0.6107
~fixed asset investment	0.6467	0.6066	0.7078	0.7342
Trade openness	0.5534	0.5697	0.6147	0.6161
~trade openness	0.7673	0.7242	0.7035	0.6988

shows that the consistency of the antecedent variables is below 0.9, whether in the context of high or low economic growth. Therefore, neither tourism nor any other individual variable can be considered as necessary for economic growth in both 2010 and 2019. This highlights the fact that economic growth is a complex and multifaceted process, which cannot be attributed to any single indicator alone. Moreover, the current research on the relationship between tourism and economic growth does not fully establish the necessary role of tourism in achieving economic development.

The NCA technique analyzes the necessity through two estimation methods: ceiling regression and ceiling envelopment. Although these two methods are applied to different types of variables (the former for continuous variables and the latter for discrete variables), the results are more robust when the two methods are used together. Therefore, this paper reports the results of the NCA analysis based on both estimation methods. The NCA technique determines the necessity of the antecedent variable by two indicators: effect value and significance. An antecedent condition is generally considered necessary if its effect value is greater than 0.1 while its p-value is less than 0.05.

Table 4. NCA results (CR indicates ceiling regression, and CE indicates ceiling envelopment).

Condition	Methods	Accuracy		Ceiling Zone		Scope		Effect Size (d)		p-Value
		2010	2019	2010	2019	2010	2019	2010	2019	2010	2019
Tourism	CR	100%	100%	0.004	0.010	0.94	0.94	0.005	0.010	0.645	0.377
	CE	100%	100%	0.009	0.014	0.94	0.94	0.009	0.015	0.481	0.244
Foreign capital	CR	99.3%	100%	0.008	0.002	0.93	0.93	0.008	0.002	0.244	0.480
	CE	100%	100%	0.008	0.004	0.93	0.93	0.008	0.004	0.278	0.413
Human resource	CR	100%	100%	0.000	0.007	0.92	0.89	0.000	0.008	1.000	0.005
	CE	100%	100%	0.000	0.014	0.92	0.89	0.000	0.016	1.000	0.001
Innovation	CR	100%	96.7%	0.000	0.067	0.93	0.94	0.000	0.071	1.000	0.000
	CE	100%	100%	0.000	0.069	0.93	0.94	0.000	0.073	1.000	0.000
Industrial structure	CR	100%	100%	0.000	0.004	0.92	0.93	0.000	0.005	0.600	0.278
	CE	100%	100%	0.000	0.009	0.92	0.93	0.000	0.009	0.597	0.102
Fixed asset investment	CR	100%	99.6%	0.001	0.003	0.96	0.95	0.001	0.003	0.945	0.683
	CE	100%	100%	0.002	0.004	0.96	0.95	0.002	0.004	0.942	0.707
Trade openness	CR	100%	100%	0.000	0.000	0.95	0.93	0.000	0.000	1.000	1.000
	CE	100%	100%	0.000	0.000	0.95	0.93	0.000	0.000	1.000	1.000

shows that the effect values of all variables are below 0.1 in both 2010 and 2019, and only the p-values of human resources and innovation are below 0.05. Therefore, all antecedent variables do not constitute necessary conditions for economic growth, and the results of necessity revealed by the NCA technique are consistent with the fsQCA approach.

Table 5. Analysis of bottleneck level.

Economic Growth	Tourism	Foreign Capital	Human Resources	Innovation	Industrial Structure	Fixed Asset Investment	Trade Openness
0	NN/NN	NN/NN	NN/NN	NN/NN	NN/NN	NN/NN	NN/NN
10	NN/0.2	NN/NN	NN/NN	NN/NN	NN/0.0	NN/NN	NN/NN
20	0.1/0.4	NN/NN	NN/NN	NN/NN	NN/0.1	NN/NN	NN/NN
30	0.2/0.6	NN/NN	NN/NN	NN/NN	NN/0.2	NN/NN	NN/NN
40	0.3/0.8	NN/NN	NN/NN	NN/NN	NN/0.4	NN/NN	NN/NN
50	0.5/1.0	NN/NN	NN/NN	NN/NN	NN/0.5	NN/NN	NN/NN
60	0.6/1.2	NN/NN	NN/NN	NN/NN	NN/0.6	NN/NN	NN/NN
70	0.7/1.4	NN/0.2	NN/NN	NN/7.1	NN/0.7	NN/NN	NN/NN
80	0.8/1.6	1.3/0.5	NN/NN	NN/17.6	NN/0.8	0.2/NN	NN/NN
90	0.9/1.8	4.0/0.8	NN/0.6	NN/28.1	NN/0.9	0.6/0.0	NN/NN
100	1.0/2.0	6.7/1.0	NN/15.1	NN/38.7	1.0/1.0	1.0/5.9	NN/NN

Note: Forward slashes before and after indicate the results of the bottleneck analyses in 2010 and 2019, respectively. “NN” indicates “not necessary”.

further reports the results of the bottlenecks analysis levels for 2010 and 2019. The results show that in 2010, 0.6% tourism specialization and no bottlenecks for any other variables were required to achieve 60% economic growth, while 1.0% tourism specialization, 6.7% foreign investment, 1.0% industrial structure, and 1.0% fixed asset investment were required to achieve 100% economic growth and no bottlenecks for any other variables. In 2019, 1.2% of tourism specialization and 0.6% of industrial structure were required to reach 60% of the economic growth level, while 2.0% of tourism specialization, 1.0% of foreign investment, 15.1% of human resources, 38.7% of innovation, 1.0% of industrial structure, and 5.9% of trade openness were required to meet 100% of the economic growth level. The results show a significant increase in the level of bottlenecks of tourism, human resources, innovation, and trade openness for economic growth from 2010 to 2019, although all antecedent variables do not constitute a necessary condition for economic growth.

4.2. Sufficiency Analysis

According to the default settings of fsQCA and standard academic criteria, the raw consistency threshold is set at 0.8, the case frequency threshold is set at 1, and the PRI consistency threshold is set at 0.7 for truth table analysis. The fsQCA model offers three solution types: complex, parsimonious, and intermediate. Intermediate approaches have garnered significant attention for their excellent accuracy and versatility [22]. Hence, solely the intermediate solutions of the fsQCA model are presented in this paper. If an antecedent variable is present in both the parsimonious and complex solutions, it is regarded as a core variable. On the other hand, if a condition only emerges in the intermediate solution, it is seen as a peripheral variable [46]. Only configurations with core conditions are analyzed in this study.

Table 6. Configurational analysis results.

Condition	2010				2019
	Economic Growth	~Economic Growth			Economic Growth			~Economic Growth
	S1	S2	S3	S4	S5	S6	S7	S8	S9	S10
Tourism	●		⊗	⊗		●		⊗	⊗	●
Foreign capital	●	⊗	⊗	⊗	●	●	●	⊗	⊗	○
Human resources	⊗	●	■		⊗	⊗		⊗	●	■
Innovation	⊗	●	●	●			●	⊗	⊗	●
Industrial structure	⊗		⊗	⊗	⊗	⊗	⊗	●		●
Fixed assets investment	●	⊗	⊗	⊗	●	●	●	⊗	⊗	⊗
Trade openness		●		■	⊗		⊗		●	⊗
Raw coverage	0.2993	0.2925	0.2884	0.2578	0.3472	0.2919	0.3053	0.3411	0.2534	0.1864
Unique coverage	0.2993	0.0527	0.0486	0.0180	0.0159	0.0312	0.0188	0.1196	0.0405	0.0193
Consistency	0.9155	0.9458	0.9446	0.9450	0.9084	0.9094	0.9232	0.9133	0.9317	0.9503
Solution coverage	0.2993	0.3592			0.3972			0.4049
Solution consistency	0.9155	0.9343			0.9036			0.9066

Note: ● means core variables are present and ⊗ means core variables are absent; ■ means periphery variables are present and ○ means periphery variables are absent.

reports the intermediate solutions of the fsQCA model. The results show that the consistency of both the overall and individual solutions is greater than 0.8, indicating the better explanatory power of each configuration for economic growth.

Table 6 shows that in 2010, there was only one solution that generated high economic growth: S1. There were three solutions that generated low economic growth: S2, S3, and S4. S1 indicates that the configuration with the core conditions of high tourism specialization, high foreign capital, low human resources, low innovation, low industrial structure, and high fixed asset investment can sufficiently lead to high economic growth. S2 shows that the configuration with the core conditions of low foreign capital, high human resources, high innovation, low fixed asset investment, and high trade openness can sufficiently lead to low economic growth. S3 shows that the configuration with low tourism specialization, low foreign capital, high innovation, low industrial structure, and low fixed asset investment as core conditions and high human resources as marginal conditions can sufficiently lead to low economic growth. S4 shows that the configuration with low tourism specialization, low foreign capital, high innovation, low industrial structure, and low fixed asset investment as core conditions can sufficiently lead to low economic growth. S3 and S4 have the same core conditions and thus have similar explanatory mechanisms for low economic growth, so they are second-order equivalence configurations. Based on the composition of each configuration, this paper develops the following model of driving or constraining economic growth in Chinese cities in 2010. Figure 2 illustrates the economic growth models’ typical cities in 2010.

The investment-led tourism-driven model corresponds to solution S1, implying that in a scenario where the level of human resources, innovation capacity, and industrial structure are low, cities promote economic growth mainly by increasing investment in fixed assets, attracting foreign investment, and developing the tourism economy. This model confirms the key role of investment in economic growth and the importance of tourism for economic growth. This is similar to the findings of existing studies on the positive impact of tourism on economic development [1,2,3,4,5,6,7,8,9]. The model also reflects a competitive commensal relationship between variables, i.e., investment and tourism growth have a crowding-out effect on the level of human resources, innovation capacity, and industrial structure optimization. In the investment-based economic growth model, the requirements for human resources and innovation capacity are not high, and because investment is mainly concentrated in infrastructure and construction industries, it drives the development of related processing and manufacturing industries and promotes the progress of secondary industries.

Although tourism is a labor-intensive industry, it is generally dominated by small and medium-sized enterprises, especially self-employment, and does not require a high overall quality of labor, so it has the same crowding-out effect on the level of human resources. In the context of China’s scenic-oriented and ticket economy tourism industry, which has become more prominent in past times, such as in 2010, the tourism industry has limited innovation capacity and role in enhancing innovation. Typical cases included in this development model include Ji’an (Jiangxi Province), Wuzhou (Guangxi Zhuang Autonomous Region), Shangrao (Jiangxi Province), Qingyuan (Guangdong Province), Yingtan (Jiangxi Province), Pingxiang (Jiangxi Province), Meishan (Sichuan Province), Jiuquan (Gansu Province), Hezhou (Guangxi Zhuang Autonomous Region), Sanmenxia (Henan Province), Jingdezhen (Jiangxi Province), and Dandong (Liaoning Province). These cities are typically characterized by increased investments in fixed assets, particularly industrial investment, and a strong focus on tourism.

The investment-constrained model corresponds to solution S2, implying that even though human resources, innovation capacity, and trade openness are at high levels, the city’s economic growth rate is low due to the constraints of foreign capital utilization and fixed asset investment. Tourism specialization does not appear in this model, i.e., this model still holds regardless of the high level of tourism development. This development model is clearly inconsistent with the endogenous economic growth theory or the new economic growth theory, which assumes that economic growth relies on technological innovation advances as well as human capital advantages. Typical examples of this model include Taiyuan (Shanxi Province), Zibo (Shandong Province), Wenzhou (Zhejiang Province), Jinhua (Zhejiang Province), Jining (Shandong Province), Taizhou (Zhejiang Province), Dongying (Shandong Province), Urumqi (Xinjiang Uygur Autonomous Region), Zhoushan (Zhejiang Province), Liaocheng (Shandong Province), and Tangshan (Hebei Province). These cities experience limitations in terms of both domestic and foreign investments, resulting in lower rates of economic growth.

The investment–industry structure tourism-constrained model includes solutions S3 and S4, indicating that even with high innovation capacity, supported by high human resources (S3) and high trade openness (S4), cities suffer from low economic growth rates due to underinvestment, lagging industrial structure, and underdeveloped tourism. These cities exhibit strong performance in terms of scientific research inputs and patent output. However, there appears to be a gap between their innovation capacity and actual productivity. Insufficient investment and underdeveloped tertiary sectors, such as tourism, have resulted in below-average urban economic performance. These cities are dominated by traditional industrial cities, but traditional industrial development has hit a bottleneck, and new economic growth points have not worked well. Typical examples of this model include Zibo (Shandong Province), Dongying (Shandong Province), Huainan (Anhui Province), Zaozhuang (Shandong Province), Pingdingshan (Henan Province), Anyang (Henan Province), Xuchang (Henan Province), Yichang (Hubei Province), Dezhou (Shandong Province), Liaocheng (Shandong Province), Tangshan (Hebei Province), Heze (Shandong Province), and Longyan (Fujian Province).

Table 6 shows that there were three configurations generating high economic growth rates in 2019: S5–S7. There were also three configurations generating low economic growth rates: S8–S10. S5 indicates that the configuration with the core conditions of high foreign capital, low human resources, low industrial structure, high fixed asset investment, and low trade openness can be sufficient to achieve high economic growth. S6 shows that the configuration with the core conditions of high tourism specialization, high foreign capital, low human resources, low industrial structure, and high fixed asset investment can sufficiently achieve high economic growth. S7 shows that the configuration with high foreign capital, high innovation, low industrial structure, and high fixed asset investment as the core conditions can sufficiently achieve high economic growth.

S8 shows that the configuration with the core conditions of low tourism specialization, low foreign investment, low human resources, low innovation, high industrial structure, and low fixed asset investment can sufficiently lead to low economic growth. S9 shows that the configuration with the core conditions of low tourism specialization, low foreign investment, high human resources, low innovation, low fixed asset investment, and high trade openness can sufficiently lead to low economic growth. Both low human resources for S8 and high human resources for S9 correspond to low economic growth, and these two configurations are similar in terms of constraints on economic growth, including investment, innovation, and tourism. Therefore, S8 and S9 are approximate second-order equivalence configurations. S10 indicates that the configuration with high tourism specialization, high innovation, high industrial structure, low fixed asset investment, and low trade openness as core conditions and low foreign capital and high human resources as peripheral conditions can sufficiently lead to low economic growth. This paper develops the following economic growth model for Chinese cities in 2019. Figure 3 illustrates the economic growth models’ typical cities in 2019.

The investment-led model aligns with solution S5, demonstrating that the city experiences improved economic growth, irrespective of the extent of tourism development, thanks to favorable investment behavior. Similarly, investment has a certain crowding-out effect on human resources, industrial structure, and trade openness. In this model, economic growth relies heavily on investment in fixed assets and the effective utilization of foreign investment. Typical examples of this model include Xuchang (Henan Province), Jingmen (Hubei Province), Baoji (Shaanxi Province), Jincheng (Shanxi Province), Suining (Sichuan Province), Suizhou (Hubei Province), Chizhou (Anhui Province), Jingdezhen (Jiangxi Province), Zunyi (Guizhou Province), Anqing (Anhui Province), and Xinyang (Henan Province).

The investment-led tourism-driven model represents solution S6, which upholds the influence of S1 on economic growth, with investment playing a dominant role and the tourism sector having a noteworthy impact, highlighting a mutually beneficial commensalism between the two sectors. Typical examples of this model include Ji’an (Jiangxi Province), Chizhou (Anhui Province), Baoji (Shaanxi Province), Jincheng (Shanxi Province), Suining (Sichuan Province), Jingdezhen (Jiangxi Province), Yichun (Jiangxi Province), Sanmenxia (Henan Province), Zunyi (Guizhou Province), Anqing (Anhui Province), Pingxiang (Jiangxi Province), and Jiujiang (Jiangxi Province). These cities demonstrate active investment, a strong focus on tourism, and a significant contribution of this industry to their economic growth. Similar to the examples covered in 2010, cities in Jiangxi Province showed improved performance under this framework.

The investment-led innovation-driven model corresponds to solution S7 and shows that mutually beneficial commensalism of investment and innovation can drive economic growth. The model confirms both the dominant role of investment and innovation in economic growth. According to the new economic growth theory, this is a sustainable development model. Typical examples of this model include Luoyang (Henan Province), Baoji (Shaanxi Province), Bengbu (Anhui Province), Xinxiang (Henan Province), Xiangyang (Hubei Province), Xiaogan (Hubei Province), Zhuzhou (Hunan Province), Anqing (Anhui Province), Xuchang (Henan Province), Jingmen (Hubei Province), Jingdezhen (Jiangxi Province), Zunyi (Guizhou Province), and Huangshi (Hubei Province). This pattern encompasses multiple cities located in the provinces of Henan and Hubei, which in part reflects the current success of these two provinces in transitioning towards an investment- and innovation-driven approach to economic growth.

The investment–innovation tourism-constrained model corresponds to solutions S8 and S9. Despite having a strong industrial structure (S8) or successful trade (S9), the city’s economic growth rate remains low due to inadequate investment and tourism development and low capacity for innovation. Typical examples of this model include Jixi (Heilongjiang Province), Baisheng (Jilin Province), Shuangyashan (Heilongjiang Province), Heihe (Heilongjiang Province), Tongchuan (Shaanxi Province), Hegang (Heilongjiang Province), Liaocheng (Shandong Province), Siping (Jilin Province), Wuwei (Gansu Province), Shanwei (Guangdong Province), Fuxin (Liaoning Province), Chifeng (Inner Mongolia Autonomous Region), Zhanjiang (Guangdong Province), and Dongying (Shandong Province). These cities are basically traditional resource or energy industry-based cities. As a result of resource depletion or shifts in energy usage, the economic growth of these urban centers is being challenged. Constrained by the conventional development paradigm, these urban areas experience impediments to innovation and tourism growth, making transformation a challenging task. Take Hegang and Jixi as examples, both of which are coal energy bases in China. However, with the depletion of coal resources and the promotion of low carbon transition in China, the economic growth of both cities has faced challenges. The limited resources and weak economy have resulted in low fixed asset investment and hindered the ability to attract foreign investment.

The investment–trade-constrained model represents solution S10, which reflects a considerable level of tourism specialization, innovation, and industrial structure, yet insufficient investment and trade constraints hinder economic growth. In this model, the contribution of tourism growth to higher economic growth rates is limited, which contradicts the traditional theoretical view that tourism fosters economic growth. Coupled with the tourism-centric economic growth model discussed earlier, this highlights the complex and unequal effects of tourism on economic development, mandating a contextual analysis of each individual case site. A typical case of this model is Lanzhou (Gansu Province). Lanzhou is an important center of higher education and scientific research in China, and its innovation capacity is ranked 80 out of 272 cities, which is a good performance. Lanzhou is also a cultural city with a long history and was selected as an excellent tourist city in China in 2004. With a tourism specialization level of 0.2702 in 2019, Lanzhou has a high economic status of tourism. However, despite the high level of innovation and tourism development, Lanzhou’s economic growth has not been positively impacted as there are noticeable deficiencies in the areas of investment and trade.

Table A1. Brief information on typical cities in 2010.

City	Province	Population Density (People per Km2)	Tourism Revenue (100 Million Yuan)	Tourist Arrivals (10,000 Person-Times)
Investment-led tourism-driven model
Ji’an	Jiangxi	190	100	1415
Wuzhou	Guangxi	229	52	665
Shangrao	Jiangxi	289	143	1987
Qingyuan	Guangdong	194	108	667
Yingtan	Jiangxi	316	45	612
Pingxiang	Jiangxi	484	48	731
Meishan	Sichuan	413	65	991
Jiuquan	Gansu	7	38	406
Hezhou	Guangxi	166	38	504
Sanmenxia	Henan	213	80	1459
Jingdezhen	Jiangxi	302	67	1325
Dandong	Liaoning	160	219	2281
Investment-constrained model
Taiyuan	Shanxi	602	230	2023
Zibo	Shandong	760	214	2561
Wenzhou	Zhejiang	754	332	3526
Jinhua	Zhejiang	490	284	2945
Jining	Shandong	722	233	3018
Taizhou	Zhejiang	594	273	3296
Dongying	Shandong	247	45	640
Urumqi	Xinjiang	226	75	782
Zhoushan	Zhejiang	769	142	2139
Liaocheng	Shandong	671	51	876
Tangshan	Hebei	534	77	1538
Investment–industry structure tourism-constrained model
Zibo	Shandong	760	214	2561
Dongying	Shandong	247	45	640
Huainan	Anhui	422	26	602
Zaozhuang	Shandong	817	61	948
Pingdingshan	Henan	622	69	898
Anyang	Henan	700	105	1401
Xuchang	Henan	862	32	608
Yichang	Hubei	191	104	1542
Dezhou	Shandong	538	47	971
Liaocheng	Shandong	671	51	876
Tangshan	Hebei	534	77	1538
Heze	Shandong	682	36	669
Longyan	Fujian	134	70	987

and

Table A2. Brief information on typical cities in 2019.

City	Province	Population Density (People per Km2)	Tourism Revenue (100 Million Yuan)	Tourist Arrivals (10,000 Person-Times)
Investment-led model
Xuchang	Henan	893	214	3197
Jingmen	Hubei	234	198	3269
Baoji	Shaanxi	208	937	12,267
Jincheng	Shanxi	250	669	7329
Suining	Sichuan	599	563	5831
Suizhou	Hubei	230	179	2835
Chizhou	Anhui	177	777	7043
Jingdezhen	Jiangxi	319	719	5523
Zunyi	Guizhou	205	2106	19,300
Anqing	Anhui	349	818	7754
Xinyang	Henan	344	282	4830
Investment-led tourism-driven model
Ji’an	Jiangxi	195	983	7790
Chizhou	Anhui	177	777	7043
Baoji	Shaanxi	208	937	12,267
Jincheng	Shanxi	250	669	7329
Suining	Sichuan	599	563	5831
Jingdezhen	Jiangxi	319	669	7329
Yichun	Jiangxi	299	910	7987
Sanmenxia	Henan	217	381	4391
Zunyi	Guizhou	205	2106	19,300
Anqing	Anhui	349	818	7754
Pingxiang	Jiangxi	507	704	5830
Jiujiang	Jiangxi	248	1161	9115
Investment-led innovation-driven model
Luoyang	Henan	454	1692	14,146
Baoji	Shaanxi	208	669	7329
Bengbu	Anhui	573	356	5122
Xinxiang	Henan	701	356	5202
Xiangyang	Hubei	288	478	6185
Xiaogan	Hubei	553	195	3022
Zhuzhou	Hunan	358	638	6466
Anqing	Anhui	349	818	7754
Xuchang	Henan	893	214	3197
Jingmen	Hubei	234	198	3269
Jingdezhen	Jiangxi	319	719	5523
Zunyi	Guizhou	205	2106	19,300
Huangshi	Hubei	539	192	2839
Investment–innovation tourism-constrained model
Jixi	Heilongjiang	75	83	1233
Baicheng	Jilin	73	95	530
Shuangyashan	Heilongjiang	62	32	934
Heihe	Heilongjiang	23	115	1313
Tongchuan	Shaanxi	201	163	2493
Hegang	Heilongjiang	67	64	669
Liaocheng	Shandong	707	220	2517
Siping	Jilin	221	26	550
Wuwei	Gansu	56	113	1957
Shanwei	Guangdong	620	173	971
Fuxin	Liaoning	169	118	1632
Chifeng	Inner Mongolia	48	421	1987
Zhanjiang	Guangdong	555	601	2845
Dongying	Shandong	264	206	2018
Investment–trade-constrained model
Lanzhou	Gansu	287	767	8211

in Appendix A present the basic information about the above representative cities.

4.3. Robustness Test

Typically, the resolution offered by fsQCA is dependent on the measurement of variables and their quantification, along with certain predetermined parameters, resulting in less persuasive outcomes from configurational analysis. To verify whether the main results are robust, this paper uses multiple approaches for robustness testing. Tourism specialization in this paper is assessed using the ratio of tourism revenue to GDP. Another measure is the ratio of tourist arrivals to local residents. Both tourism revenue and tourist arrivals are significant indicators of tourism development. Therefore, this paper introduces a novel metric to measure tourism specialization: the ratio of tourist arrivals to the local population [47,48]. An additional robustness test is based on the fsQCA method itself. This paper examines the effect of different calibration methods and parameter settings on fsQCA results. As mentioned previously, because there are no objective criteria and theoretical basis for determining the high and low thresholds for each variable, the calibration anchor points for each variable are used at 95%, 50%, and 5% quartiles. Although this practice is more common in existing studies, it is more subjective. To test the sensitivity of the configurational results to the calibration anchor points, the calibration anchor points are replaced from the fully in 95% quantile and the fully out 5% quantile to the 90% and 10% quantile, respectively, and the crossover points remain unchanged. Then, this paper refers to Du et al. [38], keeping Raw consistency unchanged and changing PRI consistency from 0.7 to 0.65.

If the results of the new configuration are not significantly different from the original results, particularly if the basic conditions of the configurations remain unaltered, then the central explanatory perspective is deemed unchanged, and the original results are robust. Of course, since the configuration analysis is based on the set theory idea, it is normal to have more configurations in the robustness test, but the new configurations should contain the original ones. The number of configurations identified may increase, particularly following the reduction of PRI consistency from 0.7 to 0.65, as the consistency requirement is eased. If the new configurations include all the original ones, the original findings are considered robust. The results of the robustness test indicate that when replacing the measure of tourism specialization, the number of configurations and core conditions for each configuration remained consistent in 2010 and 2019, with only a slight variation in peripheral conditions. When replacing the calibration anchors and PRI consistency thresholds, the number of configurations increases, but all configurations from Table 6 are still included. Therefore, it is confirmed through various tests that the core findings of this paper are robust (As the robustness test comprises several tables, the results of the examination are not presented in this article but can be obtained from the corresponding author upon request.).

5. Discussion

This paper has the following theoretical implications.

First, tourism is not a necessary condition for generating high or low economic growth rates, which complements the understanding of the conclusion that tourism has a significant impact on economic growth, as considered by existing studies. Existing studies have demonstrated that tourism positively affects economic growth or is a Granger cause for economic growth [1,2,3,4,5,6,7,8,9]. The findings presented in this paper indicate that there is no direct causation between tourism and economic growth, challenging the previously perceived limitations of single-variable analyses in past research. Tourism is not essential for economic growth, and the impact of tourism on economic growth is contingent upon a combination of various factors. No single tourism variable can be solely responsible for such growth. This indicates that tourism and other factors may have distinct or mutually supportive functions, and that tourism’s contribution to economic progress should be assessed through a holistic approach, considering the interplay of various factors. This aligns with the approach of coordinated development in sustainable development theory. As shown in configuration S1, tourism can be mutually beneficial commensalism with foreign investment and fixed asset investment to drive economic growth, or it can lead to low economic growth through a combination with innovation and industrial structure (e.g., configuration S10).

Second, the causal impact of tourism on urban economic growth is asymmetric. High levels of tourism specialization can both drive and constrain economic growth, as shown in configurations S1 and S10. The antithesis of the tourism-driven model of high economic growth does not constitute low economic growth. In this paper, no completely opposite configurations are found between high and low economic growth. Similarly, no pattern of high economic growth can be deduced from the opposites of configuration leading to low economic growth. Asymmetric causality is also reflected in equivalent causal chains of similar economic growth. For example, three configurations leading to low economic growth (see S2 to S4) were identified in 2010, and three configurations leading to low economic growth (see S8 to S10) and different but equivalent configurations leading to high economic growth (see S5 to S7) were identified in 2019.

Third, this paper identifies three models of tourism-driven or -constrained economic growth: investment-led tourism-driven, investment–industrial structure tourism-constrained, and investment–innovation tourism-constrained. Each impact model is dependent on its context, which sets it apart from the average results of standard regression analysis based on panel data. While panel data analysis does incorporate exploration of heterogeneity, this heterogeneity is primarily macroscopic and may not fully capture individual-level differences. This paper confirms the diversity of cases in tourism’s impact on economic growth. Tourism development can have a positive impact on economic growth, as evidenced by S1 and S6. However, there is also the possibility of lower economic growth rates, as indicated in S10. This study also reveals sluggish economic growth rates caused by tourism delays, as demonstrated in sections S3, S4, S8, and S9. It is important to note, however, that there is no intrinsic correlation or causality between these development models. The impact of tourism on economic growth remains highly complex and closely linked to local resource endowments, policy measures, and the economic environment.

Fourth, the pattern of tourism-driven economic growth is temporally stable. From 2010 to 2019, no new tourism-driven development patterns emerged, and tourism-driven patterns were largely consistent across time points. This confirms that after a decade of development, the influential role of tourism in China’s urban economic growth has not changed significantly. The coverage of the tourism-driven model also remains largely unchanged, changing only from 29.93% in 2010 to 29.19% in 2019. Therefore, the economic growth function of tourism in China still requires profound changes and more links with other factors to generate more diversified economic growth paths. At present, China’s tourism industry is still more in quantitative growth than qualitative change and cannot assume a leading role in driving urban economic growth.

Finally, tourism-based economic growth follows a spatial clustering pattern, with cities in close proximity showing similar patterns of economic development. This paper finds that the investment-led tourism-driven economic growth pattern mainly covers cities in Jiangxi Province, such as Ji’an, Shangrao, Yingtan, Pingxiang, and Jingdezhen in 2010, and Ji’an, Jingdezhen, Yichun, Pingxiang, and Jiujiang in 2019. This spatial agglomeration pattern may be related to the spatial agglomeration of the tourism industry. The concentration of the tourism industry in a particular area can facilitate economic growth and generate spatial spillover effects [49,50]. The same spatial agglomeration characteristics are found in the tourism-constrained model, where the investment–industrial structure tourism-constrained model in 2010 mainly covers Zibo, Dongying, Zaozhuang, Dezhou, Liaocheng, and Heze in Shandong Province, and the investment–innovation tourism-constrained model in 2019 mainly covers Jixi, Baicheng, Shuangyashan, Hehe, Hegang, Siping, Fuxin, and Chifeng in the northeastern provinces.

This paper puts forward a series of policy implications that are relevant and beneficial to stakeholders involved in the tourism industry. As a first step, the economic impact of tourism should not be limited to the tourism sector alone, especially in cities that heavily rely on tourism. It is important to identify and promote other major industries and resources that exist beyond tourism and actively seek to expand tourism’s intersecting boundaries to foster stronger partnerships with other sectors. Consequently, this will drive regional economic development and support an investment-led tourism-driven model. This model will continuously leverage foreign investment and fixed asset investment for future growth.

Moreover, policymakers need to systematically optimize the urban market environment to achieve optimal economic growth. The integration of tourism and other market factors is crucial in creating conditions that facilitate maximum economic expansion. Conversely, low economic growth can be attributed to a plethora of interconnected factors, such as inadequate market ecosystems and inefficient interactions between key market players. Therefore, policymakers need to pursue policies that foster economic growth by aligning these components of the market landscape in accordance with regional circumstances. This approach must be varied and equitable, while acknowledging spatial and temporal fluctuations.

Given the heterogeneous resource endowments across cities, realistic economic growth models must be adopted. These models should not be mechanically applied from mature models, as different cities have varying levels of development, resources, and technological advancements. Each city must have a clear understanding of its own tourism resources and keep track of any changes in internal and external environments to determine the most suitable economic growth model.

The utilization of foreign capital within the tourism industry is not particularly conspicuous within the Chinese tourism economy. Therefore, Chinese policymakers must contemplate enhancing the use of foreign capital within the tourism industry. This will increase the flow of foreign investment and improve industry contribution to the overall economy. By strengthening investment-led tourism policies, stimulating the development of diverse urban market ecosystems, and adopting realistic economic growth models, the tourism industry will become more sustainable and continue to drive regional economic development.

6. Conclusions

This paper uses fsQCA and NCA techniques based on configuration theory to explore the thesis of sustainable development in tourism and regional economic growth. It provides a comparative analysis of tourism’s role in promoting economic growth in various Chinese cities, examining both its necessity and sufficiency. The study analyzes tourism’s impact on urban economic growth in China at two significant time periods—2010 and 2019—to ascertain temporal heterogeneity and spatial disparities. The research indicates that tourism is not a prerequisite for urban economic growth, but over-dependence on the industry can significantly contribute to different economic development trajectories. The paper identifies a single investment-led tourism-driven economic growth model that has remained unchanged from 2010 to 2019. Additionally, it shows that an over-reliance on tourism can lead to a decline in economic growth, described as configuration S10.

Two tourism-constrained low economic growth models exist: investment–industry structure tourism-constrained model and investment–innovation tourism-constrained model. These two patterns indicate that it is challenging to elevate economic growth rates in instances where tourism specialization is low. Ultimately, significant fixed asset investment serves as a fundamental prerequisite for achieving elevated levels of economic growth, while insufficient fixed asset investment can result in diminished economic growth. The crucial role of fixed asset investment in economic growth is established in various instances. The role of foreign investment is similar. China’s urban economy is heavily reliant on investment, both foreign and fixed asset, to drive economic growth effectively. However, a lack of investment hinders further economic development. This phenomenon has not fundamentally changed from 2010 to 2019. In addition, tourism has an asymmetric causal impact on urban economic growth. The growth pattern of tourism-driven economies is stable over time and tends to cluster spatially. This is important to consider when analyzing the economic benefits and drawbacks of tourism for urban areas.

Future research can further enhance case analysis. Although both involve case-based empirical analysis, fsQCA analysis distinguishes itself from traditional case analysis. This paper conducts a straightforward qualitative analysis of the cases covered, providing empirical evidence for the mechanism of impact on economic growth. Nonetheless, the extensive sample fsQCA investigation lacks the depth and comprehensiveness of a single case study approach. Future research endeavors could involve further in-depth case studies to investigate diverse forms of tourism influencing or impeding economic development. In terms of dynamic analysis, this paper compares two different time periods, which may introduce some degree of randomness into the cross-sectional comparison, making it difficult to account for a continuous time series such as panel analysis. In the future, fsQCA can potentially incorporate advanced time series analysis techniques, providing a more accurate representation of dynamic shifts in economic growth within tourism-based frameworks. Regarding the scope of the configuration analysis, the article focuses only on the interrelationship between tourism and six additional variables. However, economic growth is a complex undertaking influenced by various factors, such as pollution [51], resources [52], and transportation [53]. Further investigation is required to thoroughly consider and integrate multiple variables. Finally, a network approach can be highly effective in analyzing the complex interaction of various tourism factors and their impact on economic growth.