Vulnerability of Tourist Cities’ Economic Systems Amid the COVID-19 Pandemic: System Characteristics and Formation Mechanisms—A Case Study of 46 Major Tourist Cities in China

Abstract

Research on the vulnerability of tourist cities’ economic systems during COVID-19 can offer insightful implications for tourism recovery and resilience. We built a vulnerability index of tourist cities’ economic systems based on sensitivity and responsiveness amid COVID-19. Taking 46 major tourist cities in China as a case, the vulnerability indices and main vulnerability-induced factors were analyzed using a comprehensive assessment model and a factor identification model. The results revealed several trends. First, after the pandemic emerged, the vulnerability of the economic systems of major tourist cities remained mostly moderate. Vulnerability could be further divided into four types across four city characteristics and four system characteristics. Second, sensitivity had a more pronounced effect on system vulnerability; crisis pressure and inbound tourism reliance exerted key influences on the vulnerability of tourist cities’ economic systems. Cities with high and relatively high vulnerability were subject to tourism reliance sensitivity factors and urban guarantee responsiveness factors. Third, the pandemic’s influence on tourist cities’ economic system vulnerability was mainly reflected in exogenous environmental stress vulnerability (i.e., due to external environmental stress), but was essentially endogenous structural imbalance vulnerability (i.e., due to imbalanced internal structures). Fourth, system vulnerability can be alleviated by reducing system sensitivity, improving system responsiveness, and enhancing the system’s engineering resilience and ecological resilience. This study not only offers an overview of the vulnerability characteristics of tourist cities’ economic systems amid the COVID-19 pandemic, but also highlights the formation mechanisms of vulnerability.

1. Introduction

Vulnerability is often defined as the possibility, level, or status of a system that is vulnerable to damage [1]. The term was first applied in disaster analysis. The evolution of vulnerability theory led to its introduction in the economics domain [2]. Later, the United Nations Development Program put forth the concept of “economic vulnerability”, defined as the capacity to bear damage caused by unexpected events over the course of economic development [3]. This notion has been framed as an important index to measure sustainable economic development at the national and regional level. The vulnerability of a tourist city’s economic system reflects the overall economic performance of a tourism-dependent city in response to the external environment. The level of vulnerability can forebode the possibility of a negative environment impairing the city’s economic system.

The novel coronavirus 2019 (COVID-19) pandemic has been marked as one of the most negative environmental factors with worldwide influence. With rapid propagation and widespread infection, the pandemic has posed tremendous threats to human life and health, as well as to global economic development [4]. In the tourism industry, the pandemic’s effects are considerable. According to an estimate of the United Nations World Tourism Organization (UNWTO), although global tourism experienced a 4% growth in 2021 compared to 2020, international tourist arrivals were still 72% below the pre-pandemic year of 2019. Particularly, tourism is not only a complicated industry closely associated with the national economic system, subsystems, and internal industry departments, but also is a sensitive sector easily affected by internal and external environmental factors, therefore, the disruptive effects of the pandemic are catastrophic to tourist cities that highly dependent on tourism [5,6,7]. This phenomenon can also explain why the pandemic has had a farther-reaching impact on the economic development of tourist cities compared with other types of cities, and the tourism-dependent cities have now become cities in crisis, which are especially vulnerable under the conditions of the pandemic [5]. Especially in terms of the economy, tourist cities have been suffering a great loss of revenue due to a precipitous drop in visitor numbers [5,6]. Meanwhile, a much bigger concern of the tourist city that has arisen from this is the social crisis. During the COVID-19 pandemic, the livelihoods and social well-being of individuals related to the tourism sector were seriously threatened [4]. Within the tourist city, tourism thus can play a significant factor to aggravate the impacts of a pandemic [5]. Usually, the higher the vulnerability in the tourist city, the worse will the negative impacts of the environmental factors (e.g., the pandemic) [7,8,9]. In this sense, understanding the vulnerability level of the tourist city and the main predictors of vulnerability is fundamental for the management of tourist cities in a situation of crisis. Only by doing so can the central and local government deploy appropriate policies to facilitate tourist city’s recovery in response to the COVID-19 pandemic and sustainable development in the future. Over the last few years, the tourism literature has provided significant insights into the impact of the pandemic outbreak on the tourism industry [4,5,6,8,9,10,11]. However, so far, no research has examined the vulnerability of a tourist city in the situation of a pandemic. The present research attempts to fill in this gap. Specifically, this study seeks to answer the following two questions: What are the main characteristics of tourist cities’ economic systems, and what is the formation mechanism of vulnerability during a crisis? The tourist cities in China have been employed as cases in this study. As China is one of the main developing countries within which tourism places a vital role, the research can offer implications for other tourism countries affected by the outbreak.

In this paper, we construct a vulnerability index system of tourist cities’ economic systems against the backdrop of COVID-19 to address cities’ vulnerability status and the major influencing factors on their economic systems. We further analyze system vulnerability formation mechanisms and propose effective countermeasures and managerial suggestions to reduce vulnerability. Finally, we establish a management framework for tourist cities’ economic system vulnerability amid the pandemic to minimize COVID-19′s influence on these cities, improve their normalized emergency management capabilities, boost the recovery of their tourism industries, and promote sustainable economic development.

2. Literature Review

2.1. Vulnerability of Tourist Cities’ Economic Systems

The vulnerability of a tourist city’s economic system embodies a constellation of concepts, including exposure, sensitivity, adaptability, adaptive capacity, and resilience [2,7]. As research in this vein continues to develop, scholars in various fields have generally come to agree that vulnerability comprises two key dimensions: sensitivity and adaptive capacity [12]. Adaptive capacity has been called “responsiveness” in some cases [13,14]. Therefore, in this article, vulnerability encompasses two dimensions (i.e., sensitivity and responsiveness). Sensitivity refers to a system’s ability to withstand damage in the case of internal disorder and external impact; responsiveness refers to the system’s ability to rapidly transition from a state of crisis to one of stability [15]. The root cause of tourist cities’ economic system vulnerability is the combined effect of accumulatively unreasonable characteristics (reflected as sensitivity) in the system’s internal structure and the disturbance and stress (reflected as responsiveness) imposed by external environmental changes in the tourism industry. Consequently, the system demonstrates an unsustainable development pattern [16]. The more vulnerable the system is, the less steady and resilient it is; in other words, greater vulnerability impedes the system’s ability to quickly return to its original stable state of development and renders the system less resilient to future shocks. The system’s potential for sustainable tourism development is accordingly hampered [1,7,17]. In essence, vulnerability is the opposite of resilience: the level of vulnerability can convey the level of resilience. The logical correlation between vulnerability and resilience is thus captured by a backward representation and mutual influence [18]. Research into tourist cities’ economic system vulnerability remains insufficient. Most related work has attempted to analyze the effects of natural disasters, climate change, economic crises, and other factors on tourist cities’ economic system vulnerability [19,20,21]. Many scholars have taken the crisis influence level as a baseline assessment of vulnerability [22], which informs a tourism economy’s capacity to recover from crisis [23]. As tourism economic system vulnerability gains academic interest, relevant issues have been explored in greater detail. For instance, the vulnerability of the tourism economy has been addressed in terms of environment–structure integration [24], industrial reliance [25], and other features. Commonly adopted research methods include entropy evaluation, set pair analysis, TOPSIS, and back propagation neural networks [26,27,28,29]. Studies have considered the tourism economy in addition to the urban ecological environment, society, and culture. However, research about the vulnerability of tourism economic systems in the event of major public health emergencies is lacking. Explorations of the main factors inducing vulnerability and the management implications of tourism crises also leave much to be desired.

2.2. Impact of COVID-19 Pandemic on Tourism

The COVID-19 pandemic has been deemed a public health emergency of international concern, which has presented worldwide challenges: it is not simply a health crisis, but an economic threat that has altered many businesses’ operations and individuals’ livelihoods [30]. Some researchers have linked the outbreak with a global recession [31]. In a study of the pandemic’s effects on businesses’ financial performance in Poland, significant variation was observed in industries’ sensitivity to this crisis [32]. Among the various industries, tourism is the one most affected by the pandemic [33]. The transmissibility of COVID-19 has hindered tourists’ mobility and travel plans. Meanwhile, on the supply side, numerous tourism businesses either suspended operations or closed to adhere to pandemic control measures [31]. COVID-19 has broadly shaped the personal and situational factors determining organizational and customer behavior as well [30,34]. In the last few years, the pandemic’s influence on tourism has drawn wide academic attention [35,36,37,38,39]. Research shows that the pandemic has created a volatile travel environment for national tourism systems [32]. In Europe, the tourism industry came to an abrupt halt and employees lost income mere weeks after COVID-19 was declared a pandemic [37]. Recently, an investigation on the Chinese tourism industry indicated that the control policies conducted by Chinese governments significantly mitigated the negative impacts of the pandemic [5]. Even so, the substantial negative impacts on the investment will slow down the recovery of the tourism sector. The negative impacts on tourism systems could persist for years [40]—it is currently “unclear when to expect a lasting recovery” [41]. Studies detailing the pandemic’s impacts on tourism have mainly focused on micro-level issues such as visitor arrival forecasts [42], tourists’ psychological needs and behavioral changes [43,44], industry responses to evolving pandemic conditions [45], and industry recovery measures [46]. For instance, research has suggested that tourists’ key needs shifted before, during, and in the perceived aftermath of the pandemic [33]. The supporting roles of organizations and firms are especially important for adapting to COVID-19 [35]. In particular, the pandemic is exacerbating economic and social inequalities in many developing countries, which is an agenda call for special attention in tourism research and practice [4]. Scholars have provided useful insight for tourism management during the pandemic; however, no research appears to have examined industry management from a more macro-level perspective. As indicated, tourism-dependent destinations are particularly vulnerable to the pandemic [41], and this crisis could be “a catalyst for restructuring the tourism industry markets” [45]. In such a case, exploring tourist cities’ vulnerability characteristics and the formation mechanisms of vulnerability will benefit the tourism industry’s responses to the pandemic while facilitating crisis management and industry restructuring. These developments are expected to foster long-term growth and a sustainable future for tourism in times of uncertainty.

3. Materials and Methods

3.1. Index System

Research regarding the vulnerability evaluation index system of tourist cities’ economic systems is sparse. A set of unified assessment standards has not yet been identified [7,16,18]. Findings related to the vulnerability assessment of tourist cities’ economic systems, particularly during COVID-19, are even less adequate. This article is accordingly systematic, scientific, timely, and operable. Our work addresses two prime dimensions of tourism-related vulnerability assessment (i.e., sensitivity and responsiveness [12,15,16]) and refers to relevant indices from Huang et al. [29], Sreya et al. [47], Wang et al. [48], Liang and Jie [49], Li [7], and Yin et al. [50] This paper also accounts for the correlation between vulnerability and resilience in reflecting on relevant indices proposed by Hu et al. [51], Zhang and Feng [52], and Lu et al. [53] from the perspective of tourism city resilience. The COVID-19 pandemic’s attributes are considered as well. Several new indices related to COVID-19 and regional medical quality are introduced into the initial indices (36 initial indices were screened). To ensure the scientific rigor of these index systems, we applied the Delphi method to identify suitable indicators. Ten experts (six scholars, two government officers, and two business executives) were invited to conduct two rounds of consultation to reach a consensus. Anonymous feedback was presented to the panel in each round. The first round of consultation was intended to confirm the need for indices. Specifically, experts identified key indicators based on the question “Which indicators do you think are important for the study?”. Indicators with a selection rate of less than 70% were removed, and several indicators were added based on experts’ advice. Twenty-eight indicators were retained in the end. The second round of consultation was intended to finalize the indices. Experts chose the most important indicators based on the question “Which indicators do you think are most important for the study?”. Indicators with a selection rate of less than 80% were eliminated. Twenty-two indicators remained for subsequent analysis following both rounds of expert consultation, including 11 sensitivity indicators and 11 adaptability indicators. Among them, sensitivity indices were negative, and adaptability indices were positive. Specifically, S9, S10, and S11 represented newly increased sensitivity indices related to the COVID-19 pandemic, whereas R9, R10, and R11 represented newly increased responsiveness indices related to the pandemic.

3.2. Index Data Normalization

Because indices in this assessment index system differed in their attributes and units, to eliminate the effects of dimension and magnitude, the range variation approach was used to standardize indicators. Raw data were converted into dimensionless values using Equations (1) and (2).

For a positive index (i.e., the bigger, the better):

$$y_j=\frac{x_j-\min x_j}{\max x_j-\min x_j}$$

(1)

For a negative index (i.e., the smaller, the better):

$$y_j=\frac{\max x_j-x_j}{\max x_j-\min x_j}$$

(2)

where y_j and x_j denote the standardized value and original value of index j, respectively; max x_j and min x_j denote the maximum and minimum values of index j, respectively; and 1≤ j ≤ n.

3.3. Index Weights

Index weights were determined next. To prevent experts’ subjective judgment and preferences from affecting weights determined via the analytic hierarchy process, the variation coefficient method was proposed to determine the index weighting coefficient:

$${\delta }_j=\frac{E}{\overline{x_j}}$$

(3)

$$E=\sqrt{\frac{1}{n}{\displaystyle \sum _{\mathrm{j}=1}^n{(x_{\mathrm{j}}\prime -{\overline{x}}_j)}^2}}$$

(4)

$$\overline{x_j}=\frac{1}{n}{\displaystyle \sum _{\mathrm{j}=1}^n{x}_j\prime }$$

(5)

where δ_j represents the variation coefficient of index j, E represents the mean squared error of index j, and $\overline{x_j}$ represents the mean value of index j.

The weight of index j (w_j) can be computed as follows:

$$w_j=\frac{{\delta }_j}{{\displaystyle \sum _{j=1}^{\mathrm{n}}{\delta }_j}}$$

(6)

Ultimately, the weights of the sensitivity (S) subsystem (including sensitivity indices only), responsiveness (R) subsystem (including responsiveness indices only), and vulnerability system (including both sensitivity indices and responsiveness indices) were respectively computed (see

Table 1. Indices and weights of tourist cities’ economic system vulnerability in the context of the COVID-19 pandemic.

Objective Layer	Criteria Layer	Index Layer	Definitions of Indices	Weights
				S	R	V
Vulnerability of Tourist City’s Economic System	Sensitivity (S)	S1 Percentage of total tourism revenue in GDP (%)	Regional reliance on tourism industry	0.0835	-	0.0561
		S2 Tourism growth elasticity coefficient ① (%)	Tourism’s resilience in terms of economic growth	0.0460	-	0.0309
		S3 Percentage of tourism foreign exchange receipts ② among total tourism revenue (%)	Regional tourism reliance on inbound tourism	0.1074	-	0.0722
		S4 Percentage of inbound tourist arrivals among total tourist arrivals (%)	Reliance on inbound tourism	0.1569	-	0.1055
		S5 Percentage of tertiary industrial output in GDP (%)	Reliance on tertiary industries	0.0362	-	0.0244
		S6 Contribution of tourism to residents’ income ③ (%)	Tourism’s level of influence on residents’ income	0.0878	-	0.0591
		S7 Urban registered unemployment rate (%)	City’s unemployment and employment burden	0.0452	-	0.0304
		S8 Number of traffic accidents	Number of traffic accidents in city	0.0563	-	0.0379
		S9 Ratio of confirmed cases of COVID-19 to number of doctors (%)	Pressure of pandemic on urban medical system	0.0945	-	0.0635
		S10 Ratio of imported cases of COVID-19 to number of doctors (%)	Prevention and control pressure of cases from foreign countries	0.1199	-	0.0806
		S11 Death rate of COVID-19 cases (%)	Hazard level of pandemic	0.1663	-	0.1116
	Responsiveness (R)	R1 GDP per capita (RMB)	City’s comprehensive economic strength	-	0.1077	0.0255
		R2 Local fiscal self-sufficiency rate ④ (%)	Local fiscal self-sufficiency capacity	-	0.0746	0.0342
		R3 Industrial structural diversity index ⑤	Development status and potential of regional industrial structure	-	0.0317	0.0328
		R4 Percentage of fixed asset investment in GDP (%)	Capability to use internal and external funds	-	0.0910	0.0436
		R5 Growth rate of total tourist arrivals (%)	Appeal of tourism	-	0.1018	0.0242
		R6 Growth rate of total tourism revenue (%)	Growth capacity of tourism economy	-	0.0958	0.0253
		R7 Abundance of tourism resources ⑥	Tourism resource endowment	-	0.1655	0.0151
		R8 Urban residents’ disposable income (RMB)	Urban residents’ economic vulnerability	-	0.0950	0.0265
		R9 Proportion of urban residents’ social medical insurance (%)	Urban residents’ medical insurance	-	0.0556	0.0486
		R10 Number of doctors for each 10,000 people	City’s medical level	-	0.0832	0.0202
		R11 Number of hospital beds for each 10,000 people	Opportunity for medical treatment per person	-	0.0981	0.0318

Notes: ^① Tourism growth elasticity coefficient = growth rate of total tourism revenue/GDP growth rate × 100%; refers to the percentage of economic growth driven by each 1% of tourism growth. ^② The tourism foreign exchange rate is calculated as the average exchange rate of RMB in 2019 according to China’s National Bureau of Statistics (1 USD = 6.8985 RMB). ^③ Contribution of tourism to residents’ income = tourist per capita spending on tourism/residents’ disposable income × 100%.^④ Fiscal self-sufficiency rate = general public budget income/general public budget expenditure × 100%. ^⑤ Industrial structural diversity index = ∑Xi LnX_i, where X_i refers to the percentage of the i-th industry’s added value in GDP (i = 1, 2, 3). ^⑥ Tourism resource abundance is calculated via the assignment weighting method. 5A, 4A, 3A, and 2A scenic spots in a city are allotted 5 points, 4 points, 3 points, and 2 points, respectively. The weighting of scores for all types of scenic spots is then added to indicate the city’s tourism resource abundance.

).

3.4. Comprehensive Assessment Model

The weights and normalized values of indices can be substituted into the following equation to obtain the vulnerability index, sensitivity index, and responsiveness index, respectively:

$$V/S/R={\displaystyle \sum _{j=1}^n{w}_j}\times y_j$$

(7)

where V, S, and R (0 < VSR < 1) denote the vulnerability index, sensitivity index, and responsiveness index, respectively; w_j stands for the index weight; and y_j indicates the normalized index value.

3.5. Vulnerability Factor Identification Model

To shed light on the main impact factors affecting tourist cities’ economic system vulnerability, the obstacle degree was adopted to construct a factor identification model and to elucidate this form of vulnerability:

$$P_{}=\frac{(1-y_j)\times w_j}{{\displaystyle \sum _{\mathrm{j}=1}^n(1-y_j)\times w_j}}\times 100\%$$

(8)

where P denotes the obstacle degree of each index; w_j denotes the index weight; and y_j denotes the normalized index value. By ranking the values of P, the main impact factors contributing to tourist cities’ economic system vulnerability can be identified.

3.6. Data Sources

Fifty major tourist cities in China constituted the study sample. These cities represented the 50 major tourism cities indicated in the China Tourism Statistical Yearbook, which are highly representative. By 12 April 2020, four cities—Yichang, Wuhan, Lasa, and Urumchi—had not published their official statistics. These four cities were therefore eliminated from our sample, leaving 46 tourist cities as research objects. The research data consisted of the most recent statistical data published by each city. Real-time pandemic data were also identified, reported via the official website of the National Health Commission of China, local health commissions, Baidu, NetEase, Tencent, Sina, and other sources. Trend extrapolation and the moving average method were respectively adopted to estimate missing values according to data characteristics.

4. Results

The vulnerability indices (V), sensitivity indices (S), and responsiveness indices (R) of 46 major tourist cities can be calculated using Equations (1)–(6), respectively (see

Table 2. Vulnerability indices of 46 Chinese major tourist cities’ economic systems in the context of the COVID-19 pandemic.

City Name	S	R	V	City Name	S	R	V
Beijing	0.3082	0.5635	0.3554	Jinan	0.0961	0.4573	0.2059
Tianjin	0.2742	0.3841	0.3561	Qingdao	0.1733	0.5322	0.2384
Shiazhuang	0.1422	0.3999	0.2451	Zhengzhou	0.1268	0.4655	0.2506
Qinhuangdao	0.3389	0.3699	0.4036	Luoyang	0.1886	0.4094	0.2676
Taiyuan	0.1982	0.4634	0.2812	Changsha	0.1369	0.5139	0.2157
Hohhot	0.2577	0.3466	0.3598	Zhangjiajie	0.2161	0.2907	0.3380
Shenyang	0.1562	0.4634	0.2653	Guangzhou	0.4938	0.4298	0.4958
Dalian	0.1408	0.4622	0.2442	Shenzhen	0.4639	0.3795	0.4784
Changchun	0.1605	0.3506	0.2854	Zhuahai	0.4095	0.4166	0.4250
Harbin	0.2639	0.3958	0.3475	Dongguan	0.2611	0.3005	0.3728
Shanghai	0.3543	0.4491	0.4083	Nanning	0.2226	0.4509	0.2901
Nanjing	0.1278	0.5019	0.2312	Guilin	0.3018	0.4525	0.3473
Wuxi	0.0959	0.4221	0.2517	Haikou	0.1197	0.3329	0.2610
Suzhou	0.1065	0.4592	0.2278	Sanya	0.4422	0.4045	0.4638
Hangzhou	0.1351	0.5825	0.2142	Chongqing	0.2142	0.4647	0.3240
Ningbo	0.1025	0.5081	0.2015	Chengdu	0.2149	0.5340	0.2879
Wenzhou	0.1709	0.4266	0.2670	Guiyang	0.2371	0.4810	0.3074
Hefei	0.1645	0.4933	0.2388	Kunming	0.1843	0.5416	0.2403
Huangshan	0.2789	0.3903	0.3455	Lijiang	0.1467	0.3426	0.2674
Fuzhou	0.2237	0.4717	0.2889	Xi’an	0.2257	0.5424	0.2733
Xiamen	0.2979	0.4453	0.3437	Lanzhou	0.2411	0.4604	0.3104
Quanzhou	0.2205	0.4433	0.3010	Xining	0.1110	0.4479	0.2220
Nanchang	0.1666	0.4306	0.2833	Yinchuan	0.1236	0.4017	0.2450

Source: Developed by the authors.

).

4.1. City Characteristics

According to the vulnerability index (V) of 46 cities and the mean (0.3016) and standard deviation (0.0734), all cities were divided into four vulnerability groups: (M + Std) < V < 1 reflects high vulnerability, M < V< (M + Std) reflects relatively high vulnerability, (M − Std) <V < M reflects medium vulnerability, and 0 < V< (M − Std) reflects low vulnerability. Among the chosen 46 cities, 13% were rated as having high vulnerability, 26% were rated as having relatively high vulnerability, 48% were rated as having medium vulnerability, and 13% were rated as having low vulnerability. The economic system vulnerability of these main tourist cities in China were therefore mostly moderate and displayed the following characteristics.

(1)

The high vulnerability group mostly contained coastal inbound tourist cities. These cities were further divided into two types based on their features. Three first-tier cities (Guangzhou, Shenzhen, and Shanghai) constituted the first type, representing major inbound tourist cities in China with primary international airports and important inbound ports. Three coastal cities (Sanya, Zhuhai, and Qinhuangdao) represented the second type: major inbound tourist cities that are also major coastal tourist cities in China.

(2)

The relatively high vulnerability group was concentrated in tourist cities driven by a single development mode and was again divided into two city types. The first type, resource-driven cities, held high-quality tourist resources as the driving force of economic development (e.g., Zhangjiajie, Huangshan, Guilin, Xiamen, and Beijing). For the second type, capital-driven cities, the regional economic growth model based on capital input could significantly fuel tourism economic development (e.g., Dongguan, Huhhot, Harbin, Lanzhou, Guiyang, Chongqing, and Tianjin).

(3)

The medium vulnerability group primarily involved regional central cities and traditionally competitive tourist cities. These cities were divided into two types as well. The first type covered 15 cities, most of which were capital cities and central cities with competitive advantages in the province/region (e.g., Zhengzhou, Hefei, Nanning, Fuzhou, Changchun, Nanchang, Taiyuan, Shenyang, Haikou, Shijiazhuang, Yinchuan, Kunming, Wenzhou, Wuxi, and Quanzhou). The second type covered traditional tourist cities—mostly those whose positions as tourist cities were established early alongside China’s economic reform and opening up. These cities bear a solid tourism foundation and visibility (e.g., Luoyang, Lijiang, Xi’an, Nanjing, Chengdu, Dalian, and Qingdao).

(4)

The low vulnerability group mostly featured cities that were neither at high risk from the pandemic nor were main inbound tourist destinations (“double not”). These cities also enjoyed a high level of regional economic development and an abundance of tourism resources (“double high”); sample cities include Suzhou, Xining, Changsha, Hangzhou, Jinan, and Ningbo.

4.2. System Characteristics

The mean distributions of sensitivity indices, responsiveness indices, and vulnerability indices for the above four groups appear in Figure 1.

(1)

The mean values of the sensitivity index, responsiveness index, and vulnerability index in the high vulnerability group were 0.4171, 0.4082, and 0.4458, respectively. This group of cities was thus characterized by high sensitivity and low responsiveness. As major inbound tourist cities, cities in this group are strongly reliant on tourism, especially inbound tourism. Pandemic-fueled tourism shutdowns led these cities’ tourism economic systems to be highly sensitive. Additionally, under dual pressure from the pandemic (i.e., prevention pressure from outbound cases and medical treatment pressure from domestic cases), these cities’ per capita medical resource allocation rate was relatively low. Their tourism economic systems correspondingly possessed low responsiveness. Therefore, this group’s vulnerability was closely related to the degree of pandemic hazard and the degree of tourism dependence.

(2)

The mean values of the sensitivity index, responsiveness index, and vulnerability index in the relatively high vulnerability group were 0.2627, 0.4146, and 0.3423, respectively. This group had systems with a relatively high sensitivity and relatively low responsiveness. First, for resource-driven cities, their regional economies are highly reliant on the tourism industry—hence their high sensitivity amid the pandemic. These cities also tend to rely on certain local world heritage sites (WHSs) to cultivate their core tourism competitiveness (e.g., Zhangjiajie’s Wulingyuan Scenic and Historic Interest Area WHS, Huangshan’s Mount Huangshan WHS, Guilin’s South China Karst WHS, and Xiamen’s Kulangsu WHS). Although WHSs are world-renowned tourist attractions for these cities, the resource endowment is small and involves a relatively singular type. Comprehensive tourist attractions in these cities thus remain limited. As such, these cities’ tourism economic systems displayed a low responsiveness. Moreover, even though capital-driven cities in our sample enjoyed a high level of economic development, the industrial structural diversity index (R3) of most cities was lower than the average of 1.96. The lack of high industrial structural diversity hindered the development of high responsiveness. This group’s vulnerability level was therefore closely related to the abundance of tourist resources and the degree of industrial structural diversity.

(3)

The mean values of the sensitivity index, responsiveness index, and vulnerability index in the medium vulnerability group were 0.1679, 0.4492, and 0.2639, respectively. This group featured systems with a low sensitivity and relatively high responsiveness. Capital cities or central cities in the province/region possessed several advantages, including a stronger economy, a more developed healthcare system, and a better tourism economy. They were also less affected by the pandemic because most are landlocked cities and not major ports of entry in China. These cities were hence generally more responsive and less sensitive in the context of the COVID-19 pandemic. Traditional tourism cities are often blessed with abundant tourist resources that readily draw potential tourists; these cities’ regional economies are also less reliant on the tourism industry (S1 was far smaller than the mean value of this indicator) and have a more rational industrial structure (R3 was far larger than the mean value of this indicator). Their tourism economic systems were less sensitive and more responsive as a result. In sum, this group’s vulnerability level was also tied to the abundance of tourist resources and the degree of industrial structural diversity.

(4)

The mean values of the sensitivity index, responsiveness index, and vulnerability index in the low vulnerability group were 0.1147, 0.4948, and 0.2145, respectively. This group displayed low sensitivity and high responsiveness. Their tourism economic systems were “double not and double high” (i.e., not at a high risk from the pandemic, not main inbound tourist destinations, high regional economic development, and high abundance of tourist resources), causing the systems to be less sensitive and more responsive. Moreover, as tourism is not a pillar industry for these locations, their dependence on tourism is limited. Their degree of regional industrial diversity also tended to be high, contributing to their strong economic resilience. Therefore, these cities’ tourism economic systems were characterized by a low sensitivity and high responsiveness in the face of sudden external pandemic shocks. The low vulnerability in their tourism economic systems resulted from coordinated development between the system’s internal and external structure.

4.3. Major Vulnerability Influencing Factors

The vulnerability factor identification model was used to calculate and rank the degree of obstacles posed by different indices. Five indices ranking at the top of the list of obstacles were chosen as major vulnerability influencing factors (see

Table 3. Major vulnerability influencing factors of 46 Chinese major tourist cities’ economic systems in the context of the COVID-19 pandemic.

No.	City Name	Factor 1	Factor 2	Factor 3	Factor 4	Factor 5	No.	City Name	Factor 1	Factor 2	Factor 3	Factor 4	Factor 5
1	Ningbo	S11	S4	S10	S3	S6	24	Changchun	S11	S4	S10	S3	S9
2	Jinan	S11	S4	S10	S3	S9	25	Chengdu	S4	S11	S10	S3	S9
3	Hangzhou	S11	S4	S10	S3	S9	26	Fuzhou	S11	S4	S10	S9	S1
4	Changsha	S11	S4	S10	S3	S6	27	Nanning	S11	S4	S10	S3	S6
5	Xining	S11	S4	S10	S3	S9	28	Quanzhou	S11	S4	S9	S1	S10
6	Suzhou	S11	S4	S10	S3	S9	29	Guiyang	S4	S11	S10	S3	S9
7	Nanjing	S11	S4	S10	S3	S9	30	Lanzhou	S4	S10	S3	S9	S6
8	Qingdao	S4	S11	S10	S9	S3	31	Chongqing	S4	S11	S10	S3	S6
9	Hefei	S11	S4	S10	S3	S6	32	Zhangjiajie	S11	S4	S10	S9	S6
10	Kunming	S11	S4	S10	S3	S9	33	Xiamen	S11	S4	S9	S10	S6
11	Dalian	S11	S4	S10	S3	S9	34	Huangshan	S11	S4	S10	S9	S6
12	Yinchuan	S11	S4	S10	S3	S1	35	Guilin	S11	S4	S10	S9	S3
13	Shijiazhuang	S11	S4	S10	S3	S9	36	Harbin	S4	S11	S3	S10	R4
14	Zhengzhou	S4	S11	S10	S3	S9	37	Beijing	S4	S11	S3	S6	S1
15	Wuxi	S11	S4	S10	S3	S9	38	Tianjin	S4	S11	S3	S9	S10
16	Haikou	S11	S4	S10	S3	S9	39	Huhhot	S11	S4	S3	S9	S6
17	Shenyang	S11	S4	S10	S9	S3	40	Dongguan	S11	S10	S6	S1	S4
18	Wenzhou	S11	S4	S10	S3	S6	41	Qinhuangdao	S4	S10	S3	S9	S6
19	Lijiang	S11	S4	S10	S9	S6	42	Shanghai	S11	S4	S6	S1	S9
20	Luoyang	S4	S10	S11	S3	S9	43	Zhuhai	S11	S10	S6	S1	R9
21	Xi’an	S4	S11	S10	S3	S9	44	Sanya	S11	S4	S10	R4	S3
22	Taiyuan	S11	S4	S3	S9	S6	45	Shenzhen	S11	S1	S10	R9	S6
23	Nanchang	S11	S4	S10	S3	S6	46	Guangzhou	S11	S1	R9	S9	S3

Source: Developed by the authors.

).

Although the type, obstacle degree, and number of major vulnerabilities influencing factors varied with the extent of vulnerability across cities, as much as roughly 98% of major vulnerability influencing factors emerged from the sensitivity subsystem. Sensitivity had a decisive impact on system vulnerability. Upon reviewing Table 1, the following findings were observed: the crisis pressure factor and inbound tourism reliance factor exerted key effects on the vulnerability of tourist cities’ economic systems. Cities with high vulnerability or relatively high vulnerability were also subject to the impacts of tourism reliance sensitivity factors and urban guarantee responsiveness factors.

(1)

Crisis pressure sensitivity and inbound tourism reliance sensitivity were major vulnerability influencing factors for all vulnerable cities. As displayed in Table 3, S11 (death rate of COVID-19 cases), S4 (percentage of inbound tourist arrivals among total tourist arrivals), S10 (ratio of imported cases of COVID-19 to number of doctors), S3 (percentage of tourism foreign exchange receipts among total tourism revenue), and S9 (ratio of confirmed cases of COVID-19 to number of doctors) represented shared main obstacles. First, S11, S10, and S9 were directly related to the pandemic, reflecting the degree of direct hazard caused by COVID-19 towards tourist cities and the level of pandemic-related pressure on the urban medical system. These crisis pressure sensitivity factors can influence system sensitivity and are positively correlated with vulnerability. Second, S4 and S3 each affected the inbound tourism market, mirroring cities’ extent of reliance on inbound tourism. These factors convey inbound tourism reliance sensitivity. Such attributes can inform system sensitivity and are positively correlated with system vulnerability.

(2)

The high vulnerability group and the relatively high vulnerability group were subject to impacts of tourism reliance sensitivity and urban guarantee responsiveness compared with the other two groups. These two groups of cities (from 29 to 46 in Table 3) were affected by crisis pressure sensitivity factors (S11, S10, and S9) and inbound tourism reliance sensitivity factors (S4 and S3), in addition to S6 (contribution of tourism to residents’ income), S1 (percentage of total tourism revenue in GDP), R9 (proportion of urban residents’ social medical insurance), and R4 (percentage of fixed asset investment in GDP). First, direct hazards and tourism shutdowns caused by the pandemic’s spread were national or even global in scope. The groups with high and relatively high vulnerability were thus equally affected by crisis pressure sensitivity and inbound tourism reliance sensitivity, as were the low and medium vulnerability groups. Second, S6 and S1 reflect the regional economy’s reliance on tourism; both are tourism reliance sensitivity factors that were positively correlated with sensitivity and vulnerability. R9 and R4 convey the guaranteed medical capacity for urban residents to navigate the pandemic and the guaranteed support capacity for the urban system to recover swiftly in terms of production and life. These urban guarantee responsiveness factors are conducive to improved system responsiveness; they were also negatively correlated with vulnerability.

4.4. Vulnerability Formation Mechanism

To further explore vulnerability, we combined scenario analysis with our previous results to investigate the attributes of the specific vulnerability formation mechanism. We chose two comparable cities with different vulnerability levels for scenario analysis. Qinhuangdao, a city with high vulnerability, and Xining, a city with low vulnerability, had a consistent abundance of tourism resources (109 and 102, respectively), urban scale (population of 3.1463 million and 2.3871 million, respectively), and levels of social and economic development (per capita GDP of 51,334 yuan and 57,932 yuan, respectively). These cities were therefore useful for exploring the vulnerability formation mechanism of tourist cities’ economic systems amid the pandemic.

(1)

Stage of pandemic outbreak. First, COVID-19 dealt a blow to the internal structure of the tourism economic system. The percentage of total tourism revenue in GDP (i.e., S1) and the tourism growth elasticity coefficient (i.e., S2) of Qinhuangdao were 60% and 3.42, respectively; these figures for Xining were 27% and 2.59, respectively. Qinhuangdao was thus more reliant on its local tourism industry. The internal structure of its tourism economic system faced a stronger impact. Qinhuangdao’s percentage of tourism foreign exchange receipts among total tourism revenue (i.e., S3) and the percentage of inbound tourist arrivals among total tourist arrivals (i.e., S4) were 1.71% and 0.51%, respectively; Xining’s figures in this regard were 0.48% and 0.13%, respectively. Qinhuangdao’s tourism industry thus exhibited more reliance on inbound tourism, which suffered a heavier blow from the pandemic. The city’s internal system structure was hence more disrupted by COVID-19. Second, the pandemic exerted pressure on the tourism economic system’s external environment. The most direct hazard brought by the pandemic involved 15 and 10 confirmed cases in Qinhuangdao and Xining, respectively. Worse still, there was one death in Qinhuangdao, indicating that the pandemic caused greater harm to this city. To cope with the hazards of the outbreak, both cities urgently transferred medical rescue resources, but the number of doctors per 10,000 people (i.e., R10) and the number of hospital beds per 10,000 people (i.e., R11) of Qinhuangdao were merely 31 and 60, respectively. These figures for Xining were 41 and 92, respectively. Social medical insurance coverage among urban residents in Qinhuangdao was also 10% lower than in Xining. The medical infrastructure and medical insurance system in Qinhuangdao were thus inferior to those in Xining, leading the former city to face greater pressure from the external environment.

(2)

Stage of vulnerability formation. The internal structure and external environment of the tourism economic systems in Qinhuangdao and Xining both changed due to the pandemic. On one hand, the pandemic exacerbated each system’s internal structural imbalance due to an unreasonable regional industrial structure, excessive reliance on the entire tourism industry and inbound tourism, and other factors. On the other hand, the soaring number of confirmed domestic COVID-19 cases and the growing number of imported cases amplified the external environmental pressure on these systems, which could easily stress urban medical insurance. The simultaneous effects of internal imbalance and external stress jointly altered the systems’ crisis pressure sensitivity, inbound tourism reliance sensitivity, tourism reliance sensitivity, and urban guarantee responsiveness. Both cities’ tourism economic systems’ sensitivity and responsiveness changed in kind. Consequently, two symptoms of vulnerability appeared: exogenous environmental stress vulnerability (i.e., due to stress from the external environment) and endogenous structural imbalance vulnerability (i.e., due to an imbalanced internal structure). Qinhuangdao and Xining thus demonstrated distinct vulnerability, despite their considerable social and economic development.

(3)

Stage of vulnerability coordination. To address vulnerability, Qinhuangdao and Xining can enhance their resilience by continuing to improve their public health systems and medical care systems in the short run. In the long run, both cities, especially Qinhuangdao, can reduce system sensitivity by appropriately enlarging the market share of domestic tourism and promoting the coordinated development of multiple industries in the region to respond to major vulnerability influencing factors. Their systems’ engineering resilience and ecological resilience will then constantly be enhanced, as will resilience management. System resilience will therefore rise while system vulnerability is diminished (Figure 2).

Overall, although the pandemic’s impacts on the vulnerability of tourist cities’ economic systems are reflected in exogenous environmental stress vulnerability, these influences have in fact aggravated instability and sensitivity within the tourism economic system. Vulnerability is essentially an endogenous structural imbalance vulnerability in this case. Put simply, because the tourism economic system features a typical diffusive structure with open and dependent characteristics [7], it strongly relies on external elements (e.g., geographical, social, economic, cultural, political, and other environmental features), as well as the exchange of material, energy, and information with the external environment at all times to ensure stable development [17]. These structural attributes contribute to the system’s vulnerability. That is, the tourism economic system can be easily disturbed by crisis events (e.g., the COVID-19 pandemic). Yet, such disturbances are occasional and unexpected. They therefore represent sudden external environmental variables and only temporarily affect the system (i.e., they are non-system elements). The influence effect on the system manifests as exogenous environmental stress vulnerability. Tourism economic system vulnerability is fundamentally determined by system elements, particularly endogenous structural features, which play decisive roles [19,23]. Because tourist cities mostly regard tourism as a core industry, these cities’ regional economic development has spurred long-term reliance on the tourism sector. Such industrial structural rigidity has led the tourism industry to develop rapidly, while other industries remain neglected or even inhibited to varying degrees. The risk of industrial structural imbalances within the system thus increases [10]. Tourism crises deal grave blows to the tourism industry and can weaken or even interrupt the flow of materials, capital, information, and other socioeconomic elements in the tourist city economic system [6]. Tourist cities’ tourism economic systems have thus come to rely excessively on this industry and are facing a development crisis, hence their endogenous structural imbalance vulnerability.

5. Discussion

5.1. Theoretical Implications

This study presents a multidimensional index model based on sensitivity and responsiveness to analyze the vulnerability of tourist cities’ economic systems during the COVID-19 pandemic. Taking 46 major tourist cities in China as a case, the vulnerability characteristics of the tourist cities’ economic systems were evaluated and analyzed. Further exploration of these cities’ main vulnerability-induced factors uncovered the formation mechanism of vulnerability. This effort enriches the understanding of tourism vulnerability, expands on related work, and contributes to this research area by investigating tourist destination management within the context of a pandemic [7]. To the best of our knowledge, this study is the first to highlight tourist cities’ vulnerability characteristics during a major public health emergency. The findings indicate that the vulnerability of tourist cities’ economic systems can be divided into four levels (low, moderate, relatively high, and high) across four city characteristics and four system characteristics. The economic systems of major tourist cities remained mostly moderate after the COVID-19 outbreak. Additionally, no research appears to have scrutinized the formation mechanisms of tourist cities’ vulnerability [9,11]. This paper aimed to fill this gap: sensitivity was found to have a more pronounced effect on system vulnerability, while crisis pressure and inbound tourism reliance exerted key influences on the vulnerability of tourist cities’ economic systems. Cities with high and relatively high vulnerability were subject to tourism reliance sensitivity factors and urban guarantee responsiveness factors. The results further indicated that the pandemic’s influence on tourist cities’ economic system vulnerability was mainly reflected in exogenous environmental stress vulnerability (i.e., due to external environmental stress), but was essentially endogenous structural imbalance vulnerability (i.e., due to imbalanced internal structures) [12,16]. Overall, the crisis-related vulnerability of tourist cities’ economic systems can be alleviated by reducing system sensitivity, improving system responsiveness, and enhancing the system’s engineering resilience and ecological resilience.

5.2. Managerial Implications

Although tourism cities with different vulnerability levels have different vulnerability-induced factors amid the COVID-19 pandemic, they focus on the four major vulnerability influencing factors, i.e., the crisis pressure sensitivity factors, the inbound tourism reliance sensitivity factors, the tourism reliance sensitivity factors, and the urban guarantee responsiveness factors. Therefore, addressing these four main influencing factors is the key to reducing the vulnerability of tourist cities’ economic systems. Based on the above results, a tourist city economic system resilience management framework can be offered to reduce the system’s vulnerability.

(1)

The system’s internal structural stability can be reinforced through the following efforts. First, tourist destinations’ public health systems can be improved via medical talent development, medical infrastructure upgrades, and a richer public health emergency management mechanism [54]. The urban medical insurance system should also be improved through a sound top-to-bottom institutional system, construction of a medical insurance informatization system, and establishment of a multi-level medical insurance mode complemented by commercial insurance and government medical assistance. These strategies would increase the system’s responsiveness and its resilience [31,35]. Second, to enhance the national holiday system (e.g., minor vacation system and paid vacation system), efforts can be made to upgrade tourist attraction facilities, infuse tourist products with local characteristics, and tap the domestic tourism market’s potential through multi-channel marketing [41,43]. These developments should enlarge the market share of domestic tourism and reduce cities’ dependence on inbound tourism. The integrated development mode of “tourism+” would give further play to the multiplier effect of tourism and integrate it with other industries. In particular, a series of industrial policies can be formulated to achieve macroscopic regulation of the flow of production elements, including of land, capital, and the labor force [38]. These elements will not be excessively concentrated within the tourism industry and will spread to other sectors. Multi-industry coordinated development will reduce regional economies’ overreliance on the tourism industry and inbound tourism, blunt system sensitivity, and foster system resilience.

(2)

Improving system resilience management is critical. System resilience includes engineering resilience and ecological resilience. Engineering resilience refers to the system’s swift recovery to a balanced state after experiencing an impact; ecological resilience means that the system can absorb an external shock without undergoing obvious changes and that the system then evolves into a new structure suited to the new environment [7]. It is therefore vitally important to heighten the level of resilience management. First, a normalized and routine tourism crisis management institution should be established, led by governments and with different competent tourism departments playing dominant roles [45,46]. This coordination can inspire the construction and improvement of managerial elements, including organizational staff, crisis management offices, and information management systems. Additionally, institutional norms can provide behavioral guidance and a source of authority to realize unified tourism crisis management; the leadership of resilience management would improve accordingly. Second, more attention should be devoted to enhancing cities’ emergency management capacity [55]. Particularly in the post-pandemic period, effective post-crisis tourism management is paramount for all tourist cities: adequate measures and emergency plans should be devised in response to retaliatory consumption driven by tourists’ compensatory travel behavior [56]. Such preparation can increase the system’s responsiveness and resilience in coping with retaliatory consumption [43]. Third, tourism workers should better understand crisis management to elevate enterprise management. Industrial transformation and upgrades should be pursued through integration between cultural and tourism industries. A new media cooperation mechanism should also be implemented to drive the diversification of cooperation platforms. Online tourism services should be introduced to expand value chains as well [36,51]. All in all, training should be promoted internally while cooperation is bolstered externally to encourage self-upgrades and the diverse development of tourism enterprises. These developments can collectively contribute to tourism enterprises’ resilience management and potential crisis immunity. Fourth, technological advances such as big data, 5G, virtual/augmented reality, and holographic imaging can be adopted to build “tourism + technology” competitive tourism products. Technological empowerment, innovation, and research and development related to novel tourism services and products can then proceed. New areas for tourism consumption can further be laid out in advance to generate competitive advantages. Finally, a vulnerability management framework can be designed for tourist cities’ economic systems (Figure 3).

6. Conclusions

This paper has analyzed the vulnerability of tourist cities’ economic systems in the context of the COVID-19 pandemic. Several conclusions can be drawn from the findings. First, among 46 major tourist city economic systems in China, most exhibited medium vulnerability (i.e., 13% showed high vulnerability, 26% showed relatively high vulnerability, 48% showed medium vulnerability, and 13% showed low vulnerability). The sample mainly covered coastal inbound tourism cities with high vulnerability, single element-driven tourism cities with relatively high vulnerability, regional central cities and traditionally competitive tourist cities with medium vulnerability, and “double not and double high” tourism cities with low vulnerability. Second, the multiple types of vulnerability characterizing tourist cities’ economic systems demonstrated varying degrees of sensitivity and responsiveness. System vulnerability was tied to the pandemic hazard, reliance on the tourism industry, industrial structural diversity, tourist resource abundance, and the regional social and economic development level. Third, the system’s major vulnerability influencing factors were generally derived from the sensitivity subsystem. Two sensitivity factors (crisis pressure and inbound tourism reliance) clearly informed system vulnerability. The high vulnerability group and relatively high vulnerability group were also influenced by sensitivity, tourism reliance sensitivity, and urban guarantee responsiveness. Fourth, the vulnerability formation mechanism revealed the influencing mechanism of the pandemic on the vulnerability of tourist cities’ economic system. Vulnerability applied throughout three stages: the pandemic outbreak, vulnerability formation, and vulnerability coordination. Vulnerability appeared to be due to exogenous environmental stress, but was actually attributable to endogenous structural imbalance vulnerability. In light of this finding, vulnerability can be assuaged by reducing the system’s sensitivity, improving its responsiveness, and enhancing its engineering resilience and ecological resilience to boost system resilience overall.

Thus, the tourist city economic system resilience management framework was proposed. Under this framework, targeted measures can be taken to reduce the tourism economic system’s sensitivity and improve its responsiveness, thereby enhancing the system’s internal structural stability. This outcome is a linchpin to reduce system vulnerability and is paramount to system resilience. Measures intended to lessen vulnerability—together with four aims to strengthen leadership, responsiveness, immunity, and competitiveness—will boost the system’s engineering resilience and ecological resilience. The tourism economic system can thus be equipped to quickly recover from impacts and to continue evolving by adapting to structural self-organized upgrades. The system will then be capable of maintaining balanced and sustainable development into the future.