Residents’ Perception of Tourism Impact, Participation and Support in Destinations under the COVID-19 Pandemic: The Intermediary Role of Government Trust

Abstract

Community residents’ support is one of the key factors affecting the development of tourist destinations. A clear understanding of influencing factors and internal transmission mechanisms in community residents’ support for tourism can lend to a theoretical basis and reference for tourism management departments to formulate relevant policies. It is therefore conducive to the rapid recovery and sustainable development of tourism amidst the intermittent outbreaks of COVID-19 pandemic. Taking Guilin as the site for a case study, this study examined the effect of local residents’ perceptions of risks due to the COVID-19 pandemic, benefits of tourism and residents’ participation in it, government trust, and support for tourism. In particular, the intermediary role of residents’ tourism participation and the moderating role of government trust has been explored. A total of 383 residents in Guilin City were selected as the sample for hierarchical regression analysis. The results showed that perceived risks had a significant negative impact on support for tourism, but no significant impact on resident participation. Tourism’s perceived benefits had a significant positive impact on both support for and participation in tourism, and residents’ participation played a partial mediating role in the relationship between perceived benefits of and support for tourism. Government trust played a significant moderating role in the relationship between tourism’s perceived benefits and residents’ participation, between perceived benefits and support for tourism, and between residents’ participation in and support for tourism.

1. Introduction

Tourism, which is the happiest industry, is an important carrier to improve residents’ life quality and meet people’s desire for a better life. In recent years, the COVID-19 pandemic outbreaks have continued intermittently, and tourism development faces great challenges. For residents of tourism destinations where tourism is the main industry, as well as bringing benefits to local communities, visitors also increase the risk of infection [1]. Therefore, exploring the impact mechanism of destination on residents’ risk perception of the epidemic and tourism benefit perception on tourism support amidst the intermittent outbreaks of COVID-19 pandemic will help local tourism management departments to formulate tourism development plans. It will also help them to promote the recovery and sustainable development of local tourism industries.

Residents’ support for tourism development plays an important role in improving urban competitiveness and promoting sustainable urban development [2]. The sustainable development of tourist destinations depends not only on environmental resources but also on residents’ understanding and support [3]. Community residents are part of local tourism destinations, and they are not only the creators of their local social cultures but also the major participants and beneficiaries of their tourism industries [4]. As the main population of a tourist destination, residents show support for tourism in their positive attitudes or behavioral responses to tourism.

In the field of tourism research, social exchange theory is widely used to explain the support attitude of residents to tourism development. Most studies have shown that both from the perspective of positive/negative impact perception [5] and from the perspective of economy, social culture, environment and other fields [6], all have a significant positive impact on the tourism support of residents in the destination. However, some researchers have shown that even if tourism development can significantly optimize the natural, cultural, and economic environments around tourism destinations, many residents will still be indifferent or even opposed to tourism development [7]. Many factors, such as quality of life [8], government trust [9], residents’ participation [10], tourism knowledge and power [11], and external environment [12] may be important components in the model of residents’ support for tourism development. Residents will evaluate the credibility of government agencies according to the results of government policies or actions. Government trust has a great impact on residents’ attitude, judgment and acceptance of tourism development, and is considered as an important pre-test variable of tourism support [13]. Residents’ participation in local tourism activities has a positive impact on their attitude toward tourism development [14], a critical factor that affects tourism support [15]. People tend to be more concerned about, and to show more support for, what they are involved in; they hope smooth development where their involvement is concerned. Residents engaged in tourism are more welcoming to tourists and supportive of the development of tourism [16]. Therefore, considering the mediating role of residents’ participation and the moderating role of government trust can complement the existing research results.

Based on perceived risk theory, any decision may lead to uncertain outcomes and affect the recognition and assessment of negative outcomes [17]. People instinctively avoid risk and destination residents will also take instinctual self-protection measures when facing the health and safety threat of the epidemic. However, according to the theory of social exchange [18], if perceived benefits exceed perceived costs, local residents choose to accept tourists to support tourism [19]. Studies have more often examined the driving impact on residents’ support for tourism of actual costs (including transportation noise, economic effects, etc.) and benefits brought by tourism development in tourism destinations [20]. However, there is a lack of research on the mechanism of influence for residents’ perceptions of risk as these perceptions affect willingness to support tourism during the pandemic. To sum up, this study is based on the theory of social exchange, taking Guilin, a world-famous tourist city, as a research case to explore the impact of the residents’ perception of COVID-19 pandemic risks and tourism benefits on their willingness to support tourism. The three goals expected to be achieved are: (1) To clarify the impact of epidemic risk perception and tourism benefit perception on tourism support; (2) explore the intermediary role of residents’ participation in tourism destinations; and (3) discuss the regulatory role of government trust. Compared with previous studies, this impact mechanism considers the role of residents’ participation and government trust and uses hierarchical regression analysis to comprehensively clarify the complex impact mechanism among variables, in order to provide reference for tourism destinations to formulate relevant risk management and control policies.

2. Literature Review and Research Hypotheses

The social exchange theory believes that people’s behavior towards a specific goal depends on the comparison between the cost of participating in the goal and the available realistic value. If the cost paid can be exchanged for the required value or benefit, the participation or support will be generated under the stimulation of high-quality value. If the required value or benefit is not enough to match the cost paid, It will produce negative behavior intention under the cost of “living beyond the means” [21]. The social exchange theory involves the interaction between people and resources, and both hope to obtain greater benefits at a lower cost. While the social exchange theory provides a theoretical framework for the research on tourism support, the previous research on tourism support mostly focused on the impact of material and economic benefits on the attitude and behavior of residents, ignoring the hidden costs brought by the external environment that cannot be ignored, so that the formation process of tourism development support cannot be well understood and interpreted in any situation. Therefore, exploring the impact of epidemic risk perception and tourism benefit perception on tourism support is also an extension and expansion of the social exchange theory.

The perceived impacts of tourism form a classic aspect of tourism research. Residents’ support for tourism is based on their perceptions of the positive and negative impacts of tourism [5]. Historically, the research focusses first on residents’ perceptions of economic impacts [22]. After that, it gradually began to shift toward residents’ perception of social, cultural, environmental and other non-economic impacts [23], including residents’ happiness of life [24], place image [25], community participation [26], etc. Through the development of tourism, community residents were found to perceive the positive impacts of improving their economic income, improving their living environment, and promoting cultural exchanges [27]. Meanwhile they also perceived negative impacts such as cost of goods and housing as well as noise and congestion [28]. Thus, the influence of perceived impact on residents’ support for tourism has attracted academic attention. Social Exchange Theory is the most widely used theory. The theory suggests that residents are likely to support tourism development if they perceive some personal benefits associated with tourism and believe costs will not exceed benefit [29]. Positive and negative perceptions offer the most direct antecedent variables where residents support tourism development [5]. In the context of the epidemic situation, this paper considers the epidemic risk perception as a hidden cost and discusses the impact of residents’ perception on their decision-making behavior under public crisis events, which deserves further attention.

Perceived risk theory refers to individual, subjective expectations of potential losses [30]. Perceived risk is a subjective reflection of people’s inner worlds when they face objective risks; it is also a representation of people’s willingness to accept objective risks [31]. In tourism, political instability, religious conflict, natural disaster and increasing frequency of infectious diseases pose external risks [32]. Risk exists in any decision, and human behavior decision is closely related to risk. Residents’ attitudes towards tourism, attitudes towards tourists and support for tourism are negatively affected by risk perception [33].

The development of tourist destinations depends mainly on the joint efforts of various local stakeholders, and residents play an important part. They are often integrated with the important group of participants in tourism development. Research shows that perceived impacts of tourism, as a category of residents’ perceptions, has a significant impact on residents’ participation in [34] and support for tourism [35]. These perceived impacts have a significant impact and can also reflect residents’ attitudes toward tourism development [36]. According to social exchange theory, people make reasonable choices under the comparison of positive and negative effects [18]. Positive effects promote individual decision-making and negative effects suppress it. As such, the following relationships were hypothesized.

H1: 

An epidemic’s perceived risks have significant negative impacts on residents’ participation.

H2: 

Tourism’s perceived benefits have significant positive impacts on residents’ participation.

H3: 

Epidemic’s perceived risks have significant negative impacts on residents’ support for tourism.

H4: 

Tourism’s perceived benefits have significant positive impacts on residents’ support for tourism.

Residents’ participation refers to a kind of regional public participation within the scope of a community. This participation is a process in which community residents voluntarily participate, in certain ways and forms of decision-making, management, and supervision for community affairs [37]. In community governance, community residents’ continuous participation is an important way to solve problems in community management [38]. Participation is an important link in the diversification of community governance as it mobilizes the enthusiasm of community participation and gives play to the subjectivity of community participation [39]. Residents are the main stakeholders in the development of destination tourism, and sustainable development in tourism cannot be separated from residents’ participation. Resident participation can reduce the inhibitory effect of residents’ negative perception on behavioral intention [26], at the same time, residents’ participation is an important prerequisite for support for tourism, which plays a partial mediating role in local identity and willingness to promote the development of tourism [40]. The higher degree of participation among community residents, the greater their support for tourism. Sun and Bao proposed that a high degree of participation among community residents is essential for tourism development to be truly effective and sustainable [41].

H5: 

Residents’ participation has a significant positive impact on support for tourism.

H6: 

Residents’ participation has a mediating role in the relationship between an epidemic’s perceived risks and support for tourism.

H7: 

Residents’ participation has a mediating role in the relationship between perceived benefits of and support for tourism.

Government trust mainly refers to citizens’ personal cognitions of the degree to which government policies or actions conform to their psychological expectations [42]. Studies have shown that the public’s trust in government comes from the awareness of the government’s performance in protecting the public’s safety and in preventing and controlling epidemics [43]. The negative impact of related risks on members of the public and their families can be evaluated by establishing a trust system for government agencies; after that, personal behavior can be determined [44]. It is generally believed that the perceived impact of tourism is the influencing factor of government trust, and that the perception of tourism’s benefits can enhance government trust and maintain a good relationship between the government and public [9,45].

The degree of residents’ trust in their government affects the degree of smoothness in community activities because trust leads residents to particular evaluations of government behavior and to particular attitudes of rejection. Thus, community activities may or may not find full support [46]. Government trust is the cornerstone ensuring residents’ participation, and it plays a decisive role in residents’ participation [14]. Various degrees of trust in the government lead to degrees of tolerance for the government which in turn affect residents’ views on the results of tourism development and their willingness to support tourism activities [47]. Studies have shown that government trust can significantly enhance residents’ participation and support in the process of tourism development [48]. The higher the community support for residents to participate in tourism, the higher the enthusiasm of residents to participate in the development of tourism [49]. At the same time, residents’ willingness to participate plays a role in reflecting policy [50], and their variable participation in activities may affect their political attitudes and their level of trust. Trust is seen as one of the key variables affecting the relationship between community residents and tourism development [51].

H8: 

Government trust plays a significant moderating role in the relationship between epidemics’ perceived risks and residents’ participation.

H9: 

Government trust plays a significant moderating role in the relationship between tourism’s perceived benefits and residents’ participation.

H10: 

Government trust plays a significant moderating role in the relationship between epidemics’ perceived risks and support for tourism.

H11: 

Government trust plays a significant moderating role in the relationship between tourism’s perceived benefits and support for tourism.

H12: 

Government trust plays a significant moderating role in the relationship between residents’ participation and support for tourism.

According to the above hypotheses, a conceptual model was constructed, including perceived risks of the pandemic and tourism’s perceived benefits as explanatory variables, support for tourism as explained variable, residents’ participation as mediating variable, and government trust as a moderating variable (Figure 1).

3. Methods

3.1. Study Area

Guilin City, Guangxi Zhuang Autonomous Region, is a well-known international tourist destination. Guilin is one of China’s first national historical and cultural cities and one of the first tourist cities opened to the outside world. The six major districts (Figure 2) of Guilin has many tourist attractions (Elephant Trunk Hill, East West Street, the Princess Jingjiang Residence, Two Rivers and Four Lakes, Reed Flute Cave, Guilin Museum, etc.) and a long history and culture. It is the weathervane of China’s tourism business and the permanent host place of the International Forum on Tourism Trends and Prospects held by the United Nations World Tourism Organization and the Asia Pacific Tourism Association. As tourism is the leading industry in the city, the living conditions of its residents are closely related to tourism. Studying perceived risks and benefits of tourism among local residents in Guilin, and their attitudes toward tourism in the context of the pandemic, holds significance for the subsequent development of local tourism. Guilin local residents were selected as typical, representative survey subjects.

3.2. Survey Instrument

In order to ensure content validity, the measurements used in this study were all derived from a mature scale developed by mainstream scholars. The questionnaire responses were measured using a five-point Likert scale, with “strongly disagree” at 1 and “strongly agree” at 5.

The questionnaire included two parts: One concerned demographic characteristics and the other introduced the variable scale items for measurement. The measurement of perceived risks mainly referred to the one-dimensional four-item scale used in Kim’s [52] along with Jeong and Cho’s [53] research. Perceived benefits were measured with reference to Gursoy and Rutherford’s scales [54], including seven measurements. The measurement items of government trust were based on the survey results of Nunkoo [55], Ouyang [46], and Liu [56]; six measurement items were referenced and sorted out. The measurement of support for tourism variables was based on the scales of Moghavvemi [57] et al., Sirakaya [58] et al., Stylidis [59] and others, including five measurement items.

3.3. Data Collection and Sample Characteristics

Data for this study was collected in six major districts of Guilin (Xiufeng, Yanshan, Diecai, Xiangshan, Lingui, and Qixing) from 20 May to 5 July 2022. Due to the theme of tourism, residents near tourist attractions are mainly selected for surveys, and data are collected through random interviews and household surveys. During the research, participants were informed of the research contents at the beginning of the questionnaire, then the contents were verbally restated, and participants agreed to and completed the questionnaire after being told about the confidentiality of the research. We distributed 400 questionnaires, with a recovery rate of 100%. Excluding questionnaires that presented careless, incomplete answers and others that did not meet our requirements, 383 valid questionnaires yielded an efficiency of 95.7%.

Basic descriptive analysis was undertaken using SPSS 23.0. This paper mainly used hierarchical regression for data analysis and processing. Compared with the structural equation model analysis, regression analysis will be more suitable for studying the specific mechanism of action in the model and can clearly and directly present the results; Especially for the test of regulation, hierarchical regression analysis has more extensive application than the structural equation model, which can avoid the problems of measurement error destroying statistical significance and multicollinearity. It is useful in studying the difference between multiple regression models, putting core variables at the end of a model and examining whether the influence of other factors affects the contribution of an index to a regression equation. We took resident participation and support for tourism as the outcome variables, and then put mediating variables and moderating variables, stratified, into the model for regression analysis and a moderating effect test.

The demographic characteristics of the survey sample are shown in

Table 1. Demographic overview of the sample population.

Variable	Attribute	Frequency	%	Variable	Attribute	Frequency	%
Gender	Male	192	50.1%	District	Xiufeng	68	17.8%
	Female	191	49.9%		Yanshan	75	19.6%
Age	20 years or younger	25	6.5%		Diecai	50	13.1%
	21–40 years	264	68.9%		Xiangshan	53	13.8%
	41–60 years	83	21.7%		Lingui	54	14.1%
	60 years or older	11	2.9%		Qixing	83	21.7%
Education	High school or below	152	39.7%	Career	Enterprise and institution	79	17.5%
	College or Undergraduate	212	55.4%		Professionals	89	20.6%
	Master’s degree or above	19	5%		Commercial and service industry personnel	24	23.2%
Monthly income	2000 or less	51	13.3%		Agriculture, forestry, animal husbandry and fishery production personnel	36	6.3%
	2000–4000	146	38.1%		Production and transportation-related personnel	51	9.4%
	4000–6000	128	33.4%		Student	37	13.3%
	6000–8000	30	7.8%		Other	79	9.7%
	8000 or higher	28	7.3%
Marriage	unmarried	213	55.6%
	married	170	44.4%

. On gender, males accounted for 50.1% and females accounted for 49.9%. In terms of age, considering the ability and physical reasons of residents to participate in supporting tourism, the survey sample mainly selected young and middle-aged respondents, with 90% of the respondents aged 20–60. By academic degree, 55.4% of respondents had completed college or undergraduate degrees. The sample size from each district of Guilin city was relatively balanced, each district accounting for about 15–20%. Respondents with monthly incomes of 2000–4000 yuan accounted for the highest proportion, at 38.1%; the second-highest proportion, at 33.4%, earned 4000–6000 yuan. The largest proportion of respondents were engaged in services, businesses, and professions; these respondents made up 23.2% of the sample while others engaged in agriculture, forestry, animal husbandry and fishery made up the smallest proportion, at 6.3%. The mean values of the COVID-19 pandemic’s perceived risks, tourism’s perceived benefits, residents’ participation, support for tourism, and government trust were 3.23, 3.60, 3.06, 3.42, and 3.94, respectively. The mean value of each variable being greater than 3.

4. Results

4.1. Reliability and Validity Analyses of Model

SPSS 23 was used for reliability and validity analyses; validation factor analysis used AMOS 23. The Cronbach-α coefficient, commonly used for testing consistency between items under the same study variable, was greater than 0.7. Therefore, the credibility of the questionnaire results was high, and the internal consistency of each study variable was good.

Through the factor analysis, the results showed that the scale KMO detection value and Bartlett spherical test values were 0.857 and 6602.418, respectively. Significance was 0.000, indicating a high correlation between original variables within the scale; the value was suitable for the factor analysis. Factor load for the questionnaire items should be higher than 0.5; the CR value of the latent variable should be greater than 0.7, and convergence validity requires that the AVE value of the latent variable should be greater than 0.5 [60]. First, items with factor loads significantly lower than 0.5 were removed; then, items with perceived risk at 3 and government trust at 6 were removed. Confirmatory factor analysis showed χ² = 694.248, DF = 237, CMIN/DF = 2.929 < 3, RMSEA = 0.071 < 0.08, CFI = 0.927 > 0.9, TLI = 0.907 > 0.9, IFI = 0.928 > 0.9, PCFI = 0.732 > 0.5, PNFI = 0.706 > 0.5 to fit the model and data.

Table 2. Overview of factors and items in the measurement model.

Variable and Item	Mean	Estimate	Cronbach’s α	AVE	CR
Standard		>0.5	>0.7	>0.5	>0.7
PR (Perceived risks)	3.23
Visitors have inhibited my outdoor activities	3.19	0.629	0.759	0.561	0.791
Tourists have increased the likelihood of COVID-19 infections	3.50	0.788
Visitors increased my anxiety or stress about preventing COVID-19	3.00	0.817
PB (Perceived benefits)	3.60
The development of tourism has meant Guilin’s cultural customs and ancient buildings are better protected	3.54	0.589	0.882	0.521	0.883
Tourism development has enhanced my cultural and entertainment activities	3.31	0.670
The development of tourism has improved my life and health conditions	3.45	0.739
The development of tourism has increased my knowledge and broadened my vision	3.80	0.776
The development of tourism has helped me make many friends and expand my social circle	3.55	0.763
Tourism development has increased my income	3.74	0.718
Tourism development has improved my living standard	3.82	0.777
RP (Residents’ participation)	3.06
I have provided resources and assistance for local tourism development	2.99	0.781	0.835	0.515	0.840
I fully support and cooperate with various tourism development measures	2.62	0.599
I serve visitors in various forms	3.05	0.835
I often give travel advice to the relevant authorities	3.59	0.700
I have contributed to the important decisions concerning local tourism planning	3.07	0.647
ST (Support for tourism)	3.42
I think tourism should be promoted during the COVID-19 pandemic	3.34	0.609	0.823	0.491	0.827
I support attracting more tourists to Guilin and conducting online publicity during the period of the pandemic	3.50	0.710
I support providing more effective services for tourists during the pandemic	3.60	0.668
I support the development of more tourism projects related to Guilin’s characteristic culture during the period of the pandemic	3.45	0.799
I support further investment in tourism development during the COVID-19 pandemic	3.20	0.703
GT (Government trust)	3.94
I believe in the rationality of the government’s decisions thus far	3.96	0.523	0.803	0.502	0.835
I am confident the government will give full consideration to local benefits	4.02	0.699
I believe in the decisions made by the government concerning tourism development in the context of the pandemic	4.08	0.785
Local authorities are willing to listen to residents’ tourism planning suggestions against the background of the pandemic	4.05	0.749
During the pandemic, the local competent authorities will hold meetings on tourism development	3.58	0.755

shows construct validity, expressed as aggregate validity and differentiation validity. The item factor load values for perceived risks, perceived benefits, residents’ participation and government trust were all greater than 0.5, and the combined reliability value for each latent variable was greater than 0.7. Meanwhile, the item factor load supported by tourism was 0.491. However, as shown by Fornell and Larcker [60], AVE values above 0.4 with CR above 0.6 are acceptable. Accordingly, the aggregate validity among the variables in this paper was good. Differentiation validity results showed that the correlation coefficient between any two variables was less than the square root of the AVE value of each variable itself. Within the acceptable range, this result indicated differentiation validity between variables.

4.2. Hierarchical Regression Analysis of Hypothesis Paths

A correlation analysis among variables involved in this study showed that there was a moderate and significant positive influence correlation between residents’ participation and perceived benefits (γ = 0.613, p < 0.001). There was a weak correlation between support for tourism and perceived risks, showing a negative influence (γ = −0.163, p < 0.05). There was a moderate correlation between support for tourism and perceived benefits, showing a significant positive influence (γ = 0.593, p < 0.001). The correlations among all variables laid the foundation for further study of the influence among variables.

Using residents’ participation and travel support as dependent variables, a hierarchical regression analysis was performed to verify hypotheses H1–H5. Considering that the sample’s demographics could influence the regression results, referring to the treatment of Horng [61], the seven items concerning demographic characteristics were standardized as control variables. Residents’ participation was then used as the dependent variable to build models M1–M3. Model M1 introduced the control variables, M2 added independent variable (perceived risks) to M1, and M3 added independent variable (perceived benefits) to M2. Then the regression model’s dependent variable was set as support for tourism, establishing models M4–M7. Model M4 took the control variable as the independent variable; M5 added perceived risks as the independent variable on the basis of M4. Model M6 added perceived benefits as the independent variable on the basis of M5, and M7 added resident participation as the independent variable on the basis of M6. The analysis results are shown in

Table 3. Results of the hierarchical linear regression analysis.

	Participation of Residents			Support for Tourism
	M1	M2	M3	M4	M5	M6	M7
Controlled variable
Gender	−0.105 *	−0.105 *	−0.038	−0.1	−0.097	−0.029	−0.018
Age	−0.036	−0.033	−0.073	0.03	0.017	−0.023	−0.002
Education	−0.090	−0.092	−0.069	0.015	0.023	0.046	0.066
Monthly income	0.033	0.035	0.026	−0.016	−0.026	−0.035	−0.043
District	−0.036	−0.036	−0.026	0.016	0.017	0.028	0.035
Career	−0.018	−0.018	0.006	−0.08	−0.079	−0.055	−0.057
Marriage	0.166 *	0.168 *	0.128 *	0.097	0.090	0.051	0.014
Independent variable
Perceived risks		0.024			−0.116 *
perceived benefits			0.504 ***			0.506 ***
Residents participation							0.287 ***
R2	0.060	0.060	0.295	0.035	0.048	0.285	0.343
ΔR2	0.060	0.001	0.235	0.035	0.013	0.237	0.058
F	3.397 **	2.994 **	17.376 ***	1.956	2.35 *	16.544 ***	19.451 ***
ΔF	3.397 **	0.221	124.530 ***	1.956	4.968 *	123.910 ***	32.890 ***

Note: * p < 0.05, ** p < 0.01, *** p < 0.001.

.

Residents’ participation used as the dependent variable, Table 3 shows the R² value of model M1 was 0.060, the interpretation of demographic characteristics was 6%. The F value changed significantly (p < 0.01), indicating that demographic variables had a significant effect on residents’ participation. Specifically, gender (M1, β = −0.105, p < 0.05) had a significant negative effect on residents’ participation, while marriage (M1, β = 0.166, p < 0.05) had a significant positive effect on residents’ participation. Model M2’s change in F value was not significant (p > 0.05), indicating that perceived risk had no significant effect on resident participation, so hypothesis H1 was not supported. Model M3’s change in F value was significant (p < 0.001) and the change in R² value was 0.235, indicating a significant positive effect of perceived benefits on residents’ participation (M3, β = 0.504, p < 0.001); H2 was supported.

When support for tourism was used as the dependent variable, the R² value of model M4 was 0.035, indicating the demographic characteristic variables could explain 3.5% of the change in support for tourism. The F test (F = 1.956, p > 0.05) indicated demographic characteristics had no significant effect on support for tourism. The change in F value for model M5 was significant (p < 0.05), and the R² value increased from 0.035 to 0.048, indicating a 1.3% interpretation of travel support. Specifically, perceived risk had a significant negative impact on support for tourism (M5, β = −0.116, p < 0.05), assuming H3 was supported. The change in F value for model M6 was significant (p < 0.001) and the R² value change was 0.237, meaning that the perceived benefits explained 23.7% of the model. Perceived benefits also had a significant positive impact on support for tourism (M6, β = 0.506, p < 0.001), assuming H4 was true. The change in the F value for model M7 was significant (p < 0.001), and the R² value change was 0.058, indicating that resident participation accounted for 5.8% of support for tourism. Resident participation had a significant positive impact on support for tourism (M7, β = 0.287, p < 0.001), assuming that H5 was verified.

4.3. Mediating Effects Test of Residents’ Participation

To further verify whether residents’ participation played a mediating role in the impact of their perceptions of tourism on support for tourism, the mediating effect of residents’ participation was tested using Process. Bootstrap sampling was used to set the number of repeated samplings to 5000. The running results (

Table 4. Mediating effect test results.

Relationship		Effect	SE	BootLLCI	BootULCI
Perceived risks → Residents’ participation → Support for tourism	Indirect effect	−0.011	0.029	−0.066	0.047
	Direct effect	−0.108	0.038	−0.186	−0.036
	Total effect	−0.098	0.051	−0.205	−0.035
Perceived benefits → Residents’ participation → Support for tourism	Indirect effect	0.143	0.032	0.085	0.230
	Direct effect	0.408	0.053	0.304	0.512
	Total effect	0.561	0.047	0.469	0.653

) showed indirect and direct effects of residents’ participation on the relationship between perceived risk and support for tourism at −0.011 and −0.087, respectively. The total effect was −0.098. The 95% confidence interval contained 0 (−0.066, 0.047), indicating that there was no mediating effect. Hypothesis H6 was not verified. Indirect and direct effect sizes between perceived benefits and support for tourism were 0.143 and 0.408, respectively, and the total effect was 0.551. The 95% confidence interval did not include 0 (0.085, 0.230), indicating a partial intermediary role for resident participation on the influence of perceived benefits and government trust. Hypothesis H7 was supported.

4.4. Moderating Effects Test of Government Trust

To test hypotheses H8–H12, this study used hierarchical regression analysis to examine the moderating effect of tourism practice experience, as proposed by Wen, Hou and Zhang et al. First, the independent and moderating variables were centralized; then the independent, control, and moderating variables (government trust) were put into the first layer of the independent variables of the hierarchical regression model. The basic models (M8, M10, M12, M14, M16) were constructed based on residents’ participation and support for tourism as the respective dependent variables. By multiplying perceived risks, perceived benefits, and resident participation with moderating variables for government trust, the three interaction terms were placed in the second level of independent variables, obtaining models M9, M11, M13, M15, and M17 (

Table 5. Results of the regression analysis for the moderating effect of government trust.

Dependent Variable	Participation of Residents				Support for Tourism
	M8	M9	M10	M11	M12	M13	M14	M15	M16	M17
Controlled variables
Gender	−0.099 *	−0.095 *	−0.05	−0.051	−0.091	−0.093	−0.041	−0.042	−0.031	−0.029
Age	−0.067	−0.082	−0.092	−0.072	−0.013	−0.008	−0.025	−0.013	−0.006	0.012
Education	−0.071	−0.067	−0.057	−0.064	0.042	0.041	0.048	0.043	0.06	0.042
Income	0.03	0.028	0.02	0.003	−0.031	−0.030	−0.029	−0.040	−0.034	−0.041
District	−0.028	−0.025	−0.022	−0.014	0.025	0.024	0.030	0.035	0.035	0.036
Career	−0.017	−0.022	0.001	0.000	−0.078	−0.077	−0.06	−0.060	−0.06	−0.066
Marriage	0.171 **	0.186 **	0.138 *	0.127 *	0.093	0.089	0.065	0.058	0.036	0.021
Independent variables
Perceived risk	0.029	0.019			−0.111 *	−0.108 *
perceived benefit			0.371 ***	0.344 ***			0.408 ***	0.392 ***
Residents to participate in									0.211 ***	0.205 ***
Moderating variable
Government trust	0.452 ***	0.446 ***	0.308 ***	0.336 ***	0.411 ***	0.413 ***	0.253 ***	0.270 ***	0.188 ***	0.218 ***
Interactive items
Perceived risk× government trust		0.087				−0.027
perceived benefit× government trust				0.147 ***				0.090 *
Residents participation× government trust										0.121 **
R2	0.263	0.270	0.370	0.390	0.215	0.216	0.334	0.341	0.362	0.375
ΔR2	0.263	0.007	0.370	0.020	0.215	0.001	0.334	0.008	0.362	0.013
F	14.776 ***	13.765 ***	24.322 ***	23.781 ***	11.348 ***	10.228 ***	20.762 ***	19.273 ***	21.088 ***	20.220 ***
ΔF	14.776 ***	3.699	24.322 ***	12.285 ***	11.348 ***	0.330	20.762 ***	4.248 *	21.088 ***	7.728 **

Note: * p < 0.05, ** p < 0.01, *** p < 0.001.

).

Table 5 shows that, when resident participation is the dependent variable, the interaction term of independent variables for perceived risk and government trust did not reach a significant level (M9). Hypothesis H8 was not supported. The interaction term for the independent variables of perceived benefit and government trust showed a significant positive effect on residents’ participation (M11, β = 0.147, p < 0.001). This indicated a significant moderating role in the influence of perceived benefits on residents’ participation. Hypothesis H9 was established.

When support for tourism was used as the dependent variable, the interaction terms of independent variables for perceived risk and government trust did not reach a significant level (M13). Hypothesis H10 was not supported. The interaction term of perceived benefits, residents’ participation, and moderating variables of government trust showed a significant positive impact on support for tourism (M15, β = 0.09, p < 0.05; M17, β = 0.121, p < 0.01). The F value of M15 (ΔF = 4.248, p < 0.05) and M17 (ΔF = 7.728, p < 0.01) showed significant changes. The changes in R² were 0.008 and 0.013, respectively, showing that the model was significant. It also showed that government trust had a significant moderating role on the influence of perceived benefits on support for tourism and of residents’ participation on support for tourism. Hypotheses H11 and H12 were tested.

In order to more intuitively show the moderating role of government trust, we took the mean addition and subtraction of a standard deviation of the variable as the grouping criterion to describe the relationships among tourism’s perceived benefits, support for tourism, and residents’ participation according to government trust. The results are shown in Figure 3. The solid line indicates weak government trust; the broken line indicates strong government trust. As can be seen from the figure, the slope of the broken line is greater than that of the solid line. Thus, with strong government trust, residents’ participation and perceived benefits had a slightly more positive effect on their support for tourism. Meanwhile, for residents with weak government trust, the positive impact of their participation and of perceived benefits on their support for tourism was relatively weak. This shows government trust played a moderating role in the relationships between tourism’s perceived benefits and support for tourism, between tourism’s perceived benefit and resident participation, and between residents’ participation and support for tourism.

To make the results more clear, we highlight the important causal relationship of the overall model (Figure 4). We also use the solid line to represent the hypothesis, which is supported, and the dotted line to represent the hypothesis is not supported.

5. Conclusions

The perceived risks of the pandemic had a significant negative impact on support for tourism but no significant impact on residents’ participation. The conclusion that perceived risks due to the pandemic have a significant negative impact on support for tourism is similar to the research results of Kim and Kang [62] et al. However, the conclusion that perceived risks due to the pandemic have no significant impact on residents’ participation is inconsistent with the research results of Richard [63]. Their results showed that residents’ perceptions of negative impacts significantly and negatively affected residents’ attitudes and participation. The reason for this difference in our study may be that tourism is highly developed in the destination Guilin; moreover, Guilin’s residents are highly dependent on tourism. Guilin is a famous city for tourists where the livelihoods and lives of local residents have been integrated with tourism. Furthermore, following the relatively early beginning of tourism industry in Guilin, it has withstood various trials and hardships while developing, and the residents have a certain psychological capacity for such circumstances. The effective prevention and control measures during the pandemic, along with a good situation in Guilin concerning such prevention and control, may help to strengthen residents’ psychological defense lines. Therefore, even if there are risks due to the pandemic, residents are still willing to contribute to the development of tourism; their willingness to participate is not significantly affected.

Tourism’s perceived benefits have a significant positive impact on both residents’ participation and support for tourism, and residents’ participation has a partial mediating effect in the influence of tourism’s perceived benefits on support for tourism. Perceived benefits have a significant positive impact on residents’ participation and on support for tourism, similar to the research results of Rasoolimanesh and Jaafar [29]. A small number of studies in tourism have focused primarily on residents’ participation as the dependent variable [64]. This study took residents’ participation as the driver of support for tourism, demonstrating the partial mediating role of residents’ participation in the relationship between their perceptions of benefits and their support for tourism, and verified the important role of residents’ participation in tourism support [10]. It also explored a new path for the mechanism of influence and transmission in support of tourism. The mediating role of residents’ participation has not been confirmed against perceived risks due to the pandemic and support for tourism. However, the conclusion that residents’ participation significantly and positively affects their support for tourism indicates that, despite perceived risks, resident participation can still be incentivized, further promoting their support for tourism.

Government trust plays a significant role in moderating the influences between tourism’s perceived benefits and residents’ participation and support, and between residents’ participation and support for tourism. Specifically, government trust can moderate the relationship between tourism’s perceived benefits and residents’ participation, and between perceived benefits and support for tourism [47]. Compared to community residents showing weak trust in government, residents with strong trust have a stronger, positive impact on participation in and support for tourism. This may be because trust is the basis for all social exchange activities, and government trust is the cornerstone to ensure residents’ participation [65]. Strong trust in government includes trust in government-related decisions on tourism and policies for local development, showing people’s willingness to cooperate and support the development of tourism and thus to increase their participation and support [48].

Compared to community residents with weak trust in government, those with strong trust have a greater positive impact on support for tourism. This may be because, compared with non-participant residents, those who take part in tourism or arrange activities and those who participate in making decisions about tourism or in arranging activities interact more with relevant departments, so it is easier to trust relevant government departments [66]. Thus, as residents participate in growth, their support for tourism increases significantly. The moderating effect of trust in government concerning perceived risks due to the COVID-19 pandemic, along with residents’ support for tourism, is not confirmed. This may be due to a good prevention and control situation in Guilin during the pandemic. The relevant government departments, as trust-givers in relationships of trust, show management behaviors through prevention and control measures which to some extent reduce residents’ perceptions of risk about the pandemic. This weakens the influence of government trust as seen in the research of Guo [67] et al. As a moderating variable, trust’s influence on the relationship between perception and support weakens if the trust-giver’s behavior inhibits perception.

6. Implications, Limitations and Prospects

The theoretical implications of this study is mainly reflected in two aspects. First, from the residents’ perspectives, this study explored the influence of perceived risks due to the COVID-19 pandemic, benefits due to tourism, and residents’ participation in and support for tourism. This study provides an empirical case and model to explain the impact mechanism of public health emergency risk perception on tourism support in tourist destinations. It allows deeper understanding of residents at tourism destinations given attitudes toward tourism in the context of the epidemic. Second, this study found residents’ participation in tourism and their trust in government are embedded in the influence chain of perceptions concerning tourism as these perceptions lend to support for tourism. Demonstrating the influence is helpful for an enriched body of research results concerning the complex mechanism of influence in residents’ support for tourism; such a body of results can be used as a reference for further research.

This significant findings also provide policy implications for the recovery and development in tourism. First, governments facing crises such as COVID-19 pandemic may restore incentives for residents at tourist destinations to support tourism. They can also weaken residents’ negative perceptions and pay more attention to residents’ incentives, so that perceived benefits can stimulate residents’ willing participation and support. Developing tourism while improving residents’ perceptions means dealing with negative perceptions, protecting benefits and safety, and enhancing a sense of identity. Encouraging residents to participate in decisions and implementations concerning local tourism development, enhancing their sense of gain and ownership, supports residents’ work and enhances their sense of pride and care in the industry’s development. Second, trust in government and other decision-makers can, among residents at tourism destinations, moderate residents’ participation in and support for tourism. Local governments, through effective communications with residents, improve residents’ participation. This solves the practical difficulties of local residents, showing the responsible behavior of government and other relevant policy makers, not only that they understand residents’ views and aim to meet their will. They can also establish relationships of trust with residents, informing a good social atmosphere in which residents may participate in tourism. Strengthening this relationship between residents and the government can guide residents to participate in and show support for the tourism industry. Managers should strive to improve their management activities and find the trust and dependence of residents.

There are some limitations in this study that need to be addressed in future studies. First, since social exchange theory assumes that assessing benefits and costs is an important predictive precursor to residents’ attitudes and behaviors, risk perception and benefit perception may influence each other, the intrinsic connection between perceived risks and benefits should be specifically considered; its influence on residents’ behavior can be explored in future studies. Second, while the data collected for this study was not longitudinal, residents’ perceptions of risks due to the pandemic may change over time as the scope and extent of the pandemic itself changes. And for destinations at different stages of life cycle development, residents’ perception of interests may also be different. Future studies need to conduct continuous follow-up tracking and thereby test the robustness of their models. Third, in the same place, the attitudes of residents of different occupational types and personality types may be different. Therefore, in future research, residents can be classified and analyzed for differences, or their internal characteristics (such as values, etc.) can be considered as moderators to obtain more meaningful research results.