Modelling Residents’ Perspectives of Tourism Opposition in US Counties with the Highest Historical Numbers of Reported COVID-19 Cases

Abstract

This work tests an extended theory of planned behavior model to examine residents’ behavioral intent to oppose tourism living in densely populated US counties with historically high rates of COVID-19 cases. The addition of three constructs serves as antecedents to the traditional theory of planned behavior constructs. Results revealed that passive and active opposition explained 67% of the variance in behavioral intent to oppose tourism. Of the proposed model hypotheses, 14 of the 15 were supported with oppositional attitudes toward tourism, subjective norms, and perceived behavioral control explaining 78% and 68% of the variance in passive and active behavioral intent, respectively. This paper contributes several theoretical implications (e.g., to ascertain residents’ opposition to tourism in the context of COVID-19, the current study employed TPB constructs and showed how TPB constructs effective predictors of residents’ intention to oppose tourism). The current study indicates that as the level of residents’ awareness of COVID-19 increases, they will have more negative attitudes, norms, opinions, and intentions toward tourism. Our findings will help inform destination marketing organizations in their efforts to navigate the best steps forward while balancing residents’ health and well-being with much-needed economic recovery.

1. Introduction

One need not have taken a trip in the last two years to realize fewer individuals are traveling these days. Though domestic travel may be up in some noted countries [1], the number of international travelers is still down. In looking at the United Nations World Tourism Organization [UNWTO] figures for international travelers dating back to 2012, one can notice numbers are nowhere near pre-COVID-19 statistics—1.0 billion in 2012, 1.2 billion in 2015, 1.46 billion in 2019, 381 million in 2020, and 409 million in 2021 [2,3]. In fact, according to the UNWTO’s panel of tourism experts [4], when asked when they thought their country would return to pre-pandemic (i.e., 2019) numbers, 44% claimed it would take until at least 2024; 48% said 2023, 3% said 2022, and 5% said it already had.

At the time of this writing (October 6, 2022), the virus has been responsible for 620,196,243 cases worldwide, taking the lives of 6,553,793 individuals [5]. One of the countries that estimated a slow return to pre-pandemic tourist figures was the United States. In the US alone, 96,590,395 people have contracted COVID-19, resulting in 1,061,883 deaths [5]. Unfortunately, at a time when many consider the pandemic a thing of the past, we are facing one of the most infectious and transmissible COVID-19 variants to date [6]. Arguably, many people are failing to continue to see just how dangerous the virus and its mutations are. In fact, they are ignoring the impacts of the virus since we are approaching the third year of the pandemic. In many ways, it is wishful thinking that we have ‘turned a corner,’ and administered vaccinations are increasing by the day (9,062,747,370 to be exact) [5]. Truth be told, vaccinations do not last forever, and as reports are claiming, recent variants are more resistant to vaccinations [7]. Considering this, the riskiest places to be, as COVID-19 cases continue to surge, are in densely populated areas where people have the greatest potential to be in contact with others [8]. In the US, for instance, roughly 86% of the country’s 331 million residents live in urban counties [9].

Of course, these urban counties are home to some of the most attractive tourist destinations in the US—Orlando, New York, Los Angeles, Miami to name a few [10]—which just so happen to possess some of the highest historical numbers of reported COVID-19 cases. If destinations are aiming to increase visitation, it seems logical to first gauge how aware their residents are of the virus and its impacts to gain a sense of potential animosity for or opposition of tourism, which could run counter to DMO aspirations as some recent research has demonstrated [11,12,13]. There is no better time to do this with increasing COVID-19 cases as destinations are considering relaxing their COVID-19 mitigation measures.

In response to the call of previous studies, who emphasized the shortage and/or lack of understanding of residents’ perspectives, attitudes, and intentions toward tourism and COVID-19 impacts. It is the first time this study has proposed a model that examines how residents’ awareness of COVID-19 may influence the TPB constructs and ultimately determine their behavioral intention to oppose tourism. Focusing on residents from the top 25 most-populated US cities with the highest percentage of historically reported COVID-19 cases, the aim of this work is to test a modified theory of planned behavior (TPB) model that includes three constructs of residents’ awareness of COVID-19 impacts (i.e., understanding of COVID-19, perceived vulnerability to COVID-19, and perceived severity of COVID-19) as antecedents to the traditional TPB constructs (i.e., attitudes about opposing tourism, subjective norms about opposing tourism, and perceived behavioral control in opposing tourism)—to ultimately explain residents’ behavioral intent to resist tourism in their community. Theoretical contributions of this work include: (1) extending the TPB framework in the context of the COVID-19 pandemic and residents’ opposition to tourism; and (2) testing the utility of a newly created scale of residents’ intended behavioral opposition to tourism. By considering residents’ perspectives of tourism in densely populated US counties, D.M.O.s and tourism planning organizations will be better informed to make decisions based on levels of residents’ intent to oppose tourism and those constructs that best explain such opposition.

This paper’s structure is organized as follows: First, we provided background about the relation between the awareness of COVID-19, TPB constructs, and behavioral intent to oppose tourism. Second, the research methodology is explained. Next, the research findings are presented and discussed with previous research. Finally, we provide the study conclusions, implications, limitations, and suggestions for future research.

2. Literature Review

2.1. Awareness of COVID-19’s Impact on Tourism Research

According to Rippetoe and Rogers [14], people embrace risk-preventative or protective behaviors when they perceive heightened vulnerability in particular situations. Similarly, people who believe they are at a high risk of contracting illnesses (e.g., COVID-19) tend to take precautionary actions [15]. Previous scholars [16,17] claimed that risk perceptions and/or risk awareness might determine individuals’ (tourists or residents) attitudes, perceptions, and intentions—and, in some instances, may even change their behaviors. For example, Cahyanto et al. found a positive and significant correlation between the perceived vulnerability of Ebola and travel avoidance, while the perceived severity of the Ebola factor was not a significant predictor of Americans’ travel avoidance [16].

Although several studies have focused on the relationship between COVID-19 impacts and tourists’ behavioral intentions [12,18,19,20,21,22], less is known about residents’ attitudes, intentions, and behaviors concerning tourism during the pandemic. In this study, residents’ awareness of COVID-19 includes three significant elements (perceived severity of COVID-19, perceived vulnerability of COVID-19, and understanding of COVID-19) and will be considered as predictors of the theory of planned behavior [T.B.P.] constructs. While the perceived severity of COVID-19 refers to a person’s level of concern about COVID-19, the perceived vulnerability of COVID-19 can be defined as perceptions of susceptibility to COVID-19 [15].

2.2. Theory of Planned Behavior

According to the TPB, a person’s intention to perform a specific behavior is an essential element of the model and can be explained by three constructs—attitudes perceived behavioral control, and subjective norms [23]. First, attitude refers to the way an individual assesses engaging in a particular behavior that might be either negative or positive. While perceived behavioral control is a person’s opinion of the potential challenges or obstacles involved in carrying out a certain behavior, subjective norms express people’s judgments of the social pressures in carrying out the behavior [23]. Several tourism researchers [17,20,24,25,26] have demonstrated significant relationships between the three predictor constructs and behavioral intentions and even extended the model by incorporating additional antecedents (e.g., emotional solidarity, authenticity, place attachment, destination image, community attachment, perceived risks, and perceived benefits) to behavioral intentions.

For example, Erul et al. [25] used the emotional solidarity construct to predict Turkish residents’ attitudes toward tourism development and ultimately explain their intentional support for tourism by using the three TPB antecedent constructs. The authors found that emotional solidarity dimensions significantly explained the attitudes construct, and ultimately all three TPB constructs were powerful predictors of behavioral intention. Furthermore, Liu et al. [20] examined the constructs affecting Chinese residents’ post-pandemic intentions for outbound travel. They found that while all TPB constructs and past outbound travel behavior were significant (positive) predictors of outbound travel intention, so too were the construct perceptions of COVID-19, but in a negative direction. Recently, Ojo et al. [26] found that US residents’ risk perceptions of COVID-19 and subjective norms were the most salient predictors of canceling trips during the pandemic.

2.3. Awareness of COVID-19 Impact and TPB Antecedents

Researchers have used some measures of COVID-19 awareness (e.g., perceived severity of, perceived vulnerability to, and understanding of COVID-19) to evaluate individuals’ intentions to act. For instance, Kim et al. [15] tested perceived threat (i.e., a combination of perceived severity and perceived vulnerability) as an antecedent to affective measures (i.e., hope and fear) to ultimately explain intentions to change behavior during the pandemic. Similarly, Liu et al. [20] found that Chinese residents’ perceptions of COVID-19 (e.g., perceived severity) negatively and significantly influence their post-pandemic outbound travel intention. In a similar vein, Ryu et al. [21] found that both vulnerability and severity were significant and positive predictors of residents’ risk perception. On the other hand, some studies [11,12,22] revealed a negative relationship between residents’ perceived risk of COVID-19 and their behaviors concerning tourism.

Furthermore, limited studies [17,18,27] explore the relation between COVID-19 impacts (either using risk perceptions or risk awareness) and TPB constructs. For instance, Bae and Chang [18] tested how South Korean residents’ risk perceptions (e.g., cognitive and affective) may influence TPB constructs and, ultimately, how these risk perceptions, in tandem with TPB factors, explain their behavioral intentions to engage in “untact” tourism (i.e., avoiding or minimizing contact with tourists). Similarly, Prasetyo et al. [27] used residents’ awareness of COVID-19 impacts (i.e., understanding of COVID-19, perceived vulnerability, and perceived severity) as antecedents to the traditional TPB constructs. Prasetyo et al. [27] indicated how these TPB constructs are significant predictors of intention to follow preventive measures against COVID-19 and how this intention explains residents’ preventive behaviors (e.g., actual and adapted) in the Philippines. Finally, Huang et al. [17] used the Health Belief Model (i.e., perceived vulnerability, perceived severity, and perceived benefits) in tandem with the TPB model (only considering attitudes) to predict tourists’ health risk preventative behavior. In a line with these previous studies, this study hypothesizes the following:

H₁. 

Residents’ understanding of COVID-19 will significantly (positively) predict their oppositional attitudes about tourism.

H₂. 

Residents’ understanding of COVID-19 will significantly (positively) predict their subjective norms about opposing tourism.

H₃. 

Residents’ understanding of COVID-19 will significantly (positively) predict their perceived behavioral control to oppose tourism.

H₄. 

Residents’ perceived vulnerability to COVID-19 will significantly (positively) predict their oppositional attitudes about tourism.

H₅. 

Residents’ perceived vulnerability to COVID-19 will significantly (positively) predict their subjective norms about opposing tourism.

H₆. 

Residents’ perceived vulnerability to COVID-19 will significantly (positively) predict their perceived behavioral control to oppose tourism.

H₇. 

Residents’ perceived severity of COVID-19 will significantly (positively) predict their oppositional attitudes about tourism.

H₈. 

Residents’ perceived severity of COVID-19 will significantly (positively) predict their subjective norms about opposing tourism.

H₉. 

Residents’ perceived severity of COVID-19 will significantly (positively) predict their perceived behavioral control to oppose tourism.

2.4. TPB Antecedents and Behavioral Intent to Oppose Tourism

Prior to COVID-19, several destinations struggled with over-tourism [19]. Despite a significant decline in domestic and international travel, the desire to travel remains strong, and it will rise sharply if a treatment (e.g., an effective vaccine) is discovered. Although residents are more careful and conscious about COVID-19 by wearing face masks, receiving vaccines, washing hands/using hand sanitizer, social distancing, working remotely, etc., they cannot completely insulate or protect themselves from visitors. Hence, it is normal for residents to oppose tourism, given the possibility of contracting and spreading COVID-19—even though tourism can contribute significantly to destinations’ economies. Similar to this, residents may worry about livelihoods and quality of life with the fear of accepting tourists and witnessing a repeat outbreak of COVID-19 while deciding whether to welcome tourists during and after the pandemic [19]. Previous studies [17,18,20,27] used TPB constructs to explain preventative intentional behaviors. While Huang et al. [17] used only attitude, others [18,20,27] tested all three TPB constructs to determine intentional preventative behaviors. Their results showed that TPB constructs significantly and positively predicted intentional preventative behaviors. That said, no one has considered how TPB antecedent constructs may explain residents’ behavioral intent to oppose tourism. Given this and based on the previous findings, we propose the following hypotheses (Figure 1):

H₁₀. 

Residents’ oppositional attitudes about tourism will significantly (positively) predict their behavioral intent to oppose tourism.

H₁₁. 

Residents’ subjective norms to oppose tourism will significantly (positively) predict their behavioral intent to oppose tourism.

H₁₂. 

Residents’ perceived behavioral control to oppose tourism will significantly (positively) predict their behavioral intent to oppose tourism.

3. Research Methods

3.1. Sampling and Data Collection Procedures

Residents of the 25 most-populated US counties with historically high rates of COVID-19 were chosen as the study population. Referencing the list of the 100-most populous counties in the US (based on US Census Bureau estimates) (World Population Review, 2022), we then determined the exact number of COVID-19 cases reported in each county since the beginning of the pandemic (using data from the COVID-19 Dashboard via Johns Hopkins University, 2022). Dividing the number of COVID-19 cases by population, we were able to identify the top 25 counties with the highest percentage of COVID-19 cases per population (Appendix A), which ranged from 21.0% to 35.5%. Our decision to select the top 25 counties was driven by our desire to capture counties from across the US. As reflected in Appendix A, most counties were found in states along the Atlantic Coast of the US This is not surprising given that nearly 40% of the US population can be found within this region of the country [28].

We determined one of the most efficient ways to gain access to residents of the 25 counties was to reach them via online community groups. As such, we identified and joined 380 different online community groups across the 25 counties via Facebook (i.e., Groups and Pages) and Reddit, approximately 15 groups per county. After obtaining permission to post the survey in each group from moderators, a link to the survey instrument (hosted on Qualtrics) was posted on three separate occasions roughly two weeks apart (March–May 2022). Knowing that not everyone who was a member of the groups would be from the targeted county, we asked potential participants their home zip codes. We then used that zip code to cross-check whether it was part of the county in question. Participants were offered a chance to receive one of five US $25 gift cards once they completed their submitted survey instrument.

The survey was closed five days after not receiving responses to the third posting. Nine hundred fifty-six individuals submitted a questionnaire. Of those, 213 were removed because some participants had taken less than the mean time (322 s) to submit. An additional 98 uncompleted questionnaires were removed. Finally, 115 questionnaires completed by participants residing outside of the 25 counties were also removed. Following these exclusions, the final sample size was 530 participants. Based on a confidence level of 95%, a population size of 331 million (i.e., number of US citizens), and a confidence interval of 5%, the ideal sample size for our study would be at least 385 individuals (Qualtrics, 2022). Our final sample surpassed this by 145 participants.

3.2. Measures and Data Analysis

Data were collected from participants on their responses to 49 Likert-scale items across seven constructs. The first three constructs (i.e., understanding of COVID-19, perceived vulnerability to COVID-19, and perceived severity of COVID-19) were related to awareness of COVID-19 impacts. Understanding of COVID-19 (five items), perceived vulnerability to COVID-19 (five items), and perceived severity of COVID-19 (seven items) were all adopted from Prasetyo et al. [27]. Three additional constructs pertaining to the theory of planned behavior (i.e., oppositional attitudes about tourism, subjective norms about opposing tourism, and perceived behavioral control for opposing tourism) were measured across 14 items. Six items served to measure oppositional attitudes as adapted from Erul et al. [25]. Five items were used to capture subjective norms as adapted from Erul et al. [25] and Wu et al. [29]. Three items measured perceived behavioral control as adapted from Wu et al. [29]. Finally, behavioral intent to oppose tourism was measured using 18 items adapted from Erul et al. [25], Joo et al. [12], and Litvin et al. [30]. All construct items were presented on 5-pt Likert scales of agreement (1 = strongly disagree; 5 = strongly agree).

Following the three steps of data cleaning (as mentioned above), tests of normality (e.g., skewness and kurtosis) were undertaken. Next, the 49 items were assessed for common method bias (given the singular form of data collection). Further, since the 18 items measuring participants’ behavioral intent to oppose tourism have never been utilized in prior research, we subjected the items to an exploratory factor analysis [E.F.A.]. This allowed us to determine the dimensionality of the construct. Finally, a two-step analytical sequence was performed to initially assess the measurement model (through confirmatory factor analysis or C.F.A.) for psychometrics and potentially identify any problematic items, followed by structural equation modeling [S.E.M.] to examine the structural model and test each of the 15 proposed model hypotheses. The CFA-SEM and psychometric analyses were undertaken using IBM AMOS v.26. EFA was conducted using IBM SPSS v.26.

4. Results

4.1. Sample Description

The largest percentage of respondents were between the ages of 31 and 40 (M = 34.4 years). A slight preponderance was identified as women (54.5%). The sample was slightly under ‘middle-class’ status [31], with the median annual household income ranging between USD 50,000 and USD 75,000. This falls in line with the median education level of at least an undergraduate degree. Nearly 70% of participants identified as Caucasian, with African American (8.7%), Asian American (7.9%), and Native American (7.2%) participants comprised far smaller percentages. Somewhat surprisingly, 24.9% of participants identified as Latino. Finally, respondents had lived in their current place of residence for 13.7 years (Median = 9.0 years).

4.2. Initial Data Analysis

Prior to establishing the measurement model and examining each hypothesis, we subjected the dataset to initial data analysis in the way of tests of normality and the common method bias [C.M.B.]. All skewness and kurtosis estimates were under the 2.0 threshold, as established by Kline [32]. CMB was also examined (in two ways) given data were collected solely through one method. To begin, each exogenous and endogenous variable was presented separately in unique sections on the questionnaire [33]. Next, all 49 items (across the seven model constructs) underwent an unrotated, single exploratory factor analysis [34]. No one factor explained more than 26.6% of the variance. We then proceeded with the subsequent analysis, given data were normally distributed and did not demonstrate evidence of C.M.B.

4.3. Exploratory Factor Analysis for Behavioral Intent to Oppose Tourism Items

Because the 18 items used to measure behavioral intent to oppose tourism had never been utilized in previous work, they were subjected to an EFA with principal axis factoring and varimax rotation (

Table 1. Exploratory factor analysis ^a results in behavioral intent to oppose tourism items.

Factor and Corresponding Item	Mean	Loading	% Var. Expl. b	α
Passive Opposition (P.O.) c,e	2.76		37.08	0.92
I do not plan to visit cultural attractions in my local area.	2.75	0.81
I do not plan to visit festivals/special events in my local area.	2.68	0.82
I do not plan to shop at stores frequented by tourists in my local area.	2.62	0.77
I do not plan to eat at restaurants frequented by tourists in my local area.	2.61	0.80
I do not plan to rent my property (e.g., VRBO/Airbnb, etc.) to potential visitors.	3.17	0.59
I will discourage those closest to me from renting their properties to visitors.	2.78	0.64
I will try to avoid tourists in my local area.	2.73	0.76
Active Opposition (A.O.) d,e	2.58		30.00	0.91
I will write a letter/make a phone call to influence policy related to tourism in my local area.	2.49	0.72
If presented to me, I would sign a petition concerning tourism in my local area.	2.68	0.66
If a meeting was held, I would attend to influence change regarding tourism in my local area.	2.52	0.87
If a meeting was held, I would attend to gather info re: tourism issues in my local area.	2.78	0.76
If a protest/march/rally regarding tourism in my local area was held, I would attend.	2.43	0.77
Cross-loading or low-loading items f removed following EFA
I do not plan to visit natural protected areas in my local area.
I do not plan to be helpful to tourists in my local area.
I will discourage those I know from visiting my local area.
I will oppose the promotion of tourism in my local area.
If I had to vote on a tourism-related topic, I would cast a vote in opposition.
Overall, I will not support tourism in my local area during the pandemic.

^a n = 265 randomly selected cases from overall data; KMO = 0.94; Bartlett’s test approximate χ² = 4914.79 (p < 0.001 level). ^b Total variance explained across two factors = 67.08%. ^c Eigenvalue = 4.45. ^d Eigenvalue = 3.60. ^e All items measured on 5-point Likert scales (1 = strongly disagree; 5 = strongly agree). ^f Items were removed based on double-loading onto multiple factors (with a coefficient of 0.30 or greater).

). This was done using 265 randomly selected cases from the full dataset. Following three iterations, six items were removed due to cross-loading issues or loadings lower than 0.30. The final EFA revealed a KMO estimate of 0.94 with a significant Bartlett’s test of sphericity (χ² = 4914.79, p < 0.001). Two identifiable dimensions resulted from the remaining 12 items—passive opposition (Cronbach α = 0.92) and active opposition (Cronbach α = 0.91)—explaining 67.08% of the variance in the behavioral intent to oppose tourism construct. Figure 2 shows the amended conceptual model reflecting the multidimensional nature of the oppositional construct (as depicted in splitting H₁₀, H₁₁, and H₁₂ each into two sub-hypotheses).

4.4. Measurement Model Analysis

Following a two-step sequence of analysis (i.e., CFA-SEM), the measurement model was initially established to then examine each of the 15 hypotheses pertaining to relationships between the eight model constructs [35]. Six additional items were removed from the model based on cross-loading items, error covariances, and low average variance extracted [AVE] scores [36]. The remaining 37 items in the measurement model had standardized factor loadings greater than 0.70, apart from five items (which were more than 0.50). Such estimates are considered acceptable by Hair et al. [37] (

Table 2. Measurement Model results.

Factor and Corresponding Item	Mean	β	CR	AVE
Understanding of COVID-19 (UC) a	4.23		0.85	0.59
I understand how COVID-19 is transmitted.	4.32	0.82 (N/A b)
I understand the incubation period of COVID-19.	4.12	0.76 (18.30)
I understand the symptoms of COVID-19.	4.25	0.79 (18.99)
I understand the protocol to follow if I have symptoms that may lead to COVID-19.	4.24	0.69 (16.41)
Perceived vulnerability to COVID-19 (P.V.C.) a	2.91		0.71	0.55
I think I am very vulnerable to COVID-19.	2.76	0.72 (N/A b)
I think my neighborhood is very vulnerable to COVID-19.	3.05	0.77 (8.94)
Perceived severity of COVID-19 (P.S.C.) a	3.99		0.84	0.51
I believe COVID-19 is a serious disease.	4.10	0.84 (N/A b)
I believe COVID-19 can lead to death.	4.24	0.80 (20.11)
I believe COVID-19 is more severe than many other diseases.	3.77	0.65 (15.62)
I believe COVID-19 can affect mental health.	4.00	0.65 (15.55)
I think the COVID-19 outbreak will continue for at least the next three months.	3.83	0.59 (13.90)
Oppositional attitudes about tourism (O.A.T.) a	2.62		0.96	0.82
Because of COVID…
…I do not support tourism in my local area.	2.63	0.91 (N/A b)
…I believe tourism should be actively discouraged in my local area.	2.66	0.92 (42.46)
…I opposed the development of new tourism facilities that will attract new visitors.	2.59	0.90 (33.97)
…my local area should oppose the promotion of tourism.	2.57	0.92 (35.92)
…many local tourist attractions should be closed or have reduced hours.	2.74	0.88 (32.22)
…I believe we should not have tourists in my local area currently.	2.59	0.91 (35.05)
Subjective norms about opposing tourism (S.N.T.) a	2.71		0.95	0.78
Most people who are important to me would want me to oppose tourism in my local area during the COVID-19 pandemic.	2.74	0.89 (N/A b)
Most people who are important to me think I should oppose tourism in my local area in the midst of the COVID-19 pandemic.	2.69	0.91 (36.23)
People whose opinions I value would prefer that I oppose tourism in my local area because of the COVID-19 pandemic.	2.71	0.90 (30.89)
People who influence my decisions would agree with me opposing tourism in my local area due to the COVID-19 pandemic.	2.71	0.86 (27.73)
Friends who are important to me would agree with me opposing tourism in my local area during the COVID-19 pandemic.	2.72	0.86 (27.59)
Perceived behavioral control (P.B.C.) a	3.04		0.77	0.53
I would have no difficulty at all in deciding to oppose tourism in my are aduring the COVID-19 pandemic.	2.85	0.85 (N/A b)
I am confident in making my own decision to oppose tourism in my are aduring the COVID-19 pandemic.	3.19	0.72 (15.20)
Deciding to oppose tourism in my local area due to the COVID-19 pandemic would be completely under my control.	3.08	0.59 (12.82)
Passive Opposition (P.O.) a	2.76		0.93	0.64
I do not plan to visit cultural attractions in my local area.	2.72	0.88 (N/Ab)
I do not plan to visit festivals/special events in my local area.	2.70	0.87 (28.22)
I do not plan to shop at stores frequented by tourists in my local area.	2.59	0.82 (24.87)
I do not plan to eat at restaurants frequented by tourists in my local area.	2.67	0.85 (26.44)
I do not plan to rent my property (e.g., VRBO/Airbnb, etc.) to potential visitors.	3.19	0.58 (14.72)
I will discourage those closest to me from renting their properties to visitors.	2.73	0.74 (21.05)
I will try to avoid tourists in my local area.	2.75	0.84 (26.26)
Active Opposition (A.O.) a	2.57		0.91	0.66
I will write a letter/make a phone call to influence policy related to tourism in my local area.	2.46	0.86 (N/A b)
If presented to me, I would sign a petition concerning tourism in my local area.	2.72	0.81 (23.09)
If a meeting was held, I would attend to influence change regarding tourism in my local area.	2.54	0.83 (24.12)
If a meeting was held, I would attend to gather info re: tourism issues in my local area.	2.75	0.74 (19.86)
If a protest/march/rally regarding tourism in my local area was held, I would attend.	2.39	0.81 (23.23)

^a All items measured on 5-point Likert scales (1 = strongly disagree; 5 = strongly agree). ^b In AMOS, one loading has to be fixed to 1; hence, t-value cannot be calculated for this item. All other t-values are significant (p < 0.001 level). Notes: n = 530, χ² = 1394.47, df = 594, χ²/df = 2.35, CFI = 0.95, TLI = 0.95, IFI = 0.95, RMSEA = 0.05.

).

CFA results revealed strong incremental model fit indices—comparative fit index (CFI), Tucker-Lewis Index (TLI), and incremental model fit index (I.F.I.)—each at 0.95 [37]. The absolute model fit (i.e., root mean square error of approximation or RMSEA), was deemed good, coming in less than 0.06 [37]. Overall, the CFA model fit was χ²(n = 530) = 1394.47, df = 594, χ²/df = 2.35, CFI = 0.95, TLI = 0.95, IFI = 0.95, RMSEA = 0.05.

The psychometric properties of the measurement model were then examined. Only two of the eight constructs displayed composite reliabilities less than 0.80 (i.e., 0.71 and 0.77). All AVE estimates were in excess of 0.50, ranging from 0.51 to 0.82. Convergent and discriminant validities were also examined to best gauge construct validity. Discriminant validity was determined based on the square roots of each construct’s AVE being greater than the inter-construct correlations [38] (

Table 3. Discriminant validity analysis results.

Factors	CR	AVE	PO	UC	PVC	PSC	OAT	SNT	PBC	PBC
Passive opposition (P.O.)	0.93	0.64	0.80
Understanding COVID-19 (UC)	0.85	0.59	0.22	0.77
Perceived vulnerability COVID-19 (P.V.C.)	0.71	0.55	0.37	0.00	0.74
Perceived severity COVID-19 (P.S.C.)	0.84	0.51	0.17	0.61	0.39	0.71
Oppositional attitudes about tourism (O.A.T.)	0.96	0.82	0.77	0.29	0.31	0.07	0.91
Subjective norms opposing tourism (S.N.T.)	0.95	0.78	0.77	0.26	0.29	0.05	0.87	0.88
Perceived behavioral control (P.B.C.)	0.77	0.53	0.58	0.02	0.24	0.22	0.58	0.62	0.73
Active opposition (A.O.)	0.91	0.66	0.74	0.39	0.23	0.12	0.80	0.79	0.46	0.81

Note: The bold diagonal elements are the square root of the variance shared between the factors and their measures (average variance extracted). Off-diagonal elements are the correlations between factors. For discriminant validity, the diagonal elements should be larger than any other corresponding row or column entry. CR: composite reliability; AVE: average variance extracted.

). In only one instance of the 28 possibilities was an inter-construct correlation (e.g., the correlation between passive behavioral intention to oppose tourism and active behavioral intention to oppose tourism) in excess of the square root of a construct’s AVE (e.g., passive behavioral intention to oppose tourism). Given this occurred within the same overarching construct, it was deemed a minor problem. With standardized factor loadings were in excess of 0.50, AVEs were greater than 0.50, and t values associated with each item standardized factor loading were significant—convergent validity was also established [37].

4.5. Structural Model Analysis

With the measurement model established, the structural model was run to examine the 15 hypotheses from the amended conceptual model. The model demonstrated good fit to the data: χ²(n = 530) = 1585.08, df = 604, χ²/df = 2.62; CFI = 0.94; TLI = 0.93; IFI = 0.94; RMSEA = 0.06. As

Table 4. Hypothesized relationships between constructs from the structural model.

Hypothesized Relationship	B	Beta (β)	t-Statistic	Supported?
H1: UC → OAT	0.39	0.21	2.42 *	Yes
H2: UC → SNT	0.43	0.25	2.55 **	Yes
H3: UC → PBC	0.32	0.19	2.71 **	Yes
H4: PSC→OAT	0.77	0.33	3.36 ***	Yes
H5: P.V.C. → SNT	0.80	0.37	3.33 ***	Yes
H6: P.V.C. → PBC	0.54	0.26	3.28 ***	Yes
H7: P.S.C. → OAT	0.55	0.36	2.89 **	Yes
H8: P.S.C. → SNT	0.59	0.42	2.96 **	Yes
H9: P.S.C. → PBC	0.38	0.28	2.79 **	Yes
H10a: O.A.T. → PO	0.76	0.81	12.84 ***	Yes
H10b: O.A.T. → AO	0.43	0.48	6.91 ***	Yes
H11a: S.N.T. → PO	0.00	0.00	0.03 ns	No
H11b: S.N.T. → AO	0.41	0.43	5.76 ***	Yes
H12a: P.B.C. → PO	0.11	0.11	2.86 **	Yes
H12b: P.B.C. → AO	0.10	0.10	2.22 *	Yes

Note: The fit indices are: χ² = 1585.08, RMSEA = 0.05, IFI = 0.94, TLI = 0.93, and CFI = 0.94. * p < 0.05, ** p < 0.01, *** p < 0.001; ns > 0.05. R²_SMC: Oppositional attitudes about tourism = 0.83, Subjective norms in opposition = 0.93, Perceived behavioral control = 0.47, Passive opposition = 0.78, and active opposition = 0.68.

indicates, 14 of the 15 hypotheses were supported. The first nine hypotheses involved residents’ awareness of COVID-19’s impact. Each of the first three hypotheses was significant—residents’ understanding of COVID-19 explained oppositional attitudes about tourism (H₁: β = 0.21, p < 0.05), subjective norms (H₂: β = 0.25, p < 0.01), and perceived behavioral control (H₃: β = 0.19, p < 0.01). The next set of hypotheses was also significant—perceived vulnerability to COVID-19 explained oppositional attitudes about tourism (H₄: β = 0.33, p < 0.001), subjective norms (H₅: β = 0.37, p < 0.001), and perceived behavioral control (H₆: β = 0.26, p < 0.001). Finally, the third set of hypotheses was significant—perceived severity of COVID-19 predicted oppositional attitudes about tourism (H₇: β = 0.36, p < 0.01), subjective norms (H₈: β = 0.42, p < 0.01), and perceived behavioral control (H₉: β = 0.28, p < 0.01). Overall, residents’ understanding of COVID-19, perceived vulnerability to COVID-19, and perceived severity of COVID-19 explained 83% of the variance in oppositional attitudes about tourism, 93% of the variance in subjective norms, and 47% of the variance in perceived behavioral control.

The remaining six hypotheses pertained to the theory of planned behavior constructs as predictors of both passive behavioral intent to oppose tourism and active behavioral intent to oppose tourism. Oppositional attitudes about tourism significantly predicted passive behavioral intent to oppose tourism (H_10a: β = 0.81, p < 0.001) and active behavioral intent to oppose tourism (H_10b: β = 0.48, p < 0.001). Subjective norms did not significantly explain passive behavioral intent to oppose tourism (H_11a: β = 0.00, p > 0.05) though they did significantly predict active behavioral intent to oppose tourism (H_11b: β = 0.43, p < 0.001). Finally, perceived behavioral control significantly explained both passive behavioral intent to oppose tourism (H_12a: β = 0.11, p < 0.01) and active intent to oppose tourism (H_12b: β = 0.10, p < 0.05). Together, oppositional attitudes about tourism, subjective norms, and perceived behavioral control explained 78% of the variance in passive behavioral intent to oppose tourism and 68% in active behavioral intent to oppose tourism.

5. Discussion

This study examined how residents’ awareness of COVID-19 (i.e., understanding of COVID-19, perceived vulnerability to COVID-19, perceived severity of COVID-19) influence TPB constructs (i.e., attitudes about opposing tourism, subjective norms about opposing tourism, and perceived behavioral control in opposing tourism) and ultimately explain their behavioral intention to oppose tourism. The first three hypotheses (H₁–H₃) showed that there were positive relationships between understanding of COVID-19 and TPB constructs. In other words, the more residents’ understand the seriousness of COVID-19, the more they oppose tourism. In contrast to our findings, Prasetyo et al. [27] found that a strong understanding of COVID-19 was not a significant predictor of attitude. However, similar to our findings (i.e., H₂ and H₃) they found that it significantly predicted subjective norms and perceived behavioral control.

In response to H₄–H₉, results indicate both perceived vulnerability to COVID-19 and perceived severity of COVID-19 were significant predictors of TPB constructs. In other words, those who showed a higher susceptibility to contracting COVID-19 and agreed with its severity also showed a higher tendency to oppose tourism. Based on those findings, it can be said that people may take preventive measures to mitigate their risk of contracting a disease (e.g., COVID-19). Those results (H₄–H₉) were similar to Prasetyo et al. [27] findings that revealed each construct (i.e., perceived vulnerability and perceived severity) significantly predicted TPB constructs. However, in contrast to our results, Huang et al. [17] found that perceived severity did not considerably influence attitude toward preventative behavior.

H10–H12 showed that TPB constructs were significant determinants of behavioral intention to oppose tourism which is consistent with previous findings [17,18,20,24,27]. Among these TPB constructs, residents’ oppositional attitudes about tourism were the strongest predictors of behavioral intentions to oppose tourism. This is in line with previous findings [18,20], however, in contrast to other scholars’ [24,27] findings which indicated residents’ perceived behavioral control was the best predictor of behavioral intention. Based on our findings, people who have oppositional attitudes about tourism are surrounded by social pressures to oppose tourism, and enough knowledge and opinions about opposing tourism are most likely to contribute to such opposition.

Briefly, the current study makes not only several theoretical contributions (i.e., an extension of the TPB framework to account for residents’ hostility to tourism in the COVID-19 epidemic and an evaluation of the usefulness of a newly developed scale measuring residents’ intentional behavioral opposition to tourism) but also practical contributions; for example, DMOs and tourism planning organizations will be better equipped to make judgments based on levels of residents’ desire to oppose tourism and those constructs that best explain such opposition by taking into account residents’ viewpoints of tourism in densely populated US counties.

6. Study Implications

This research was a response to the call by Ryu et al. [21] that emphasized the shortage and/or lack of understanding of residents’ perspectives, attitudes, and intentions toward tourism and COVID-19 impacts. As such, this is the first study that proposed a model to examine how residents’ awareness of COVID-19 may influence the TPB constructs and ultimately determine their behavioral intention to oppose tourism. Hence, this study has several theoretical contributions. First, it indicates that the factors of residents’ awareness of COVID-19 play a crucial role in determining residents’ oppositional attitudes toward tourism and ultimately explain their behavioral intentions to oppose tourism. Second, the current study applied TPB constructs to assess residents’ opposition to tourism in the context of COVID-19. Finally, this study extended the TPB model by including residents’ awareness of COVID-19 and showing how TPB constructs powerful determinants of residents’ intention to oppose tourism.

The perceived health risk (e.g., COVID-19) is one of the most essential factors in individuals’ (e.g., tourists and residents) current travel decision-making [17]. More specifically, individuals’ attitudinal, intentional and actual behavior can be influenced by their perception of their vulnerability to and severity of health risks [15,16]. The current study verified these claims by indicating that as the level of residents’ awareness of COVID-19 increases, so too do their opposing attitudes about tourism (in light of subjective norms and perceived behavioral control) which signals significant behavioral intentions to oppose tourism in their communities.

This work has implications for practice as well, especially considering densely populated destinations with high percentages of historic COVID-19 cases. As destinations welcome back more visitors, policymakers and destination managers should promote the continued use of strategies to minimize the spread of the virus that will protect both residents and tourists. Some of these strategies include highly encouraging up-to-date vaccinations (based on the U.S. Centers for Disease Control guidelines), mandating face masks/handwashing/sanitizing hands, postponing or canceling highly crowded events, limiting services (based on the percentage of rooms occupied in lodging facilities and guests allowed in restaurants and bars), encouraging touchless services, and maintaining social distances. Residents can manage their contact and interactions with visitors best when they are kept abreast of the current impacts of COVID-19. Thus, tourism authorities should inform residents about the effects of COVID-19 on tourism and explain the current state of tourism (e.g., the number of tourists, the impacts on tourism, the seriousness of COVID-19, etc.) and/or the use of technologies (e.g., robots in hospitality and travel sectors) to help them mitigate the spread of COVID-19. For example, from upon arrival to departure, destinations should encourage minimal contact with tourists using intelligent technologies (i.e., check-in, passport control, front-desk assistance, navigation, or other services). Such strategies may help put residents at ease and facilitate greater support for tourism in their communities.

Several actions can assist in risk management and risk perception reduction. To help locals feel more at ease welcoming tourists or not opposing tourism, tourism groups or managers should take initiatives to provide messaging that lowers these risk perceptions (i.e., perceived vulnerability to and perceived severity of COVID-19). For instance, government representatives should inspire public confidence and guarantee that locals will live in a safe environment by ensuring that all safety precautions are performed, tourists are screened as best as possible before they arrive at the destination, and that residents have updated vaccinations. At that point, government representatives can also utilize social media’s full potential to promote safety and a trustworthy atmosphere to lessen these risk perceptions (e.g., how well the local medical and health systems are doing to prevent the spread of infectious diseases). Similarly, tourism officials can better convey to residents and tourists the destination’s response plans should the number of cases increase. Additionally, tourism authorities should work in tandem with local and state authorities that have COVID-19 contact tracing in place.

7. Limitations and Future Research Opportunities

This research is not without limitations. Prior to this study, the items used to measure behavioral intent to oppose tourism were untested. The final KMO estimate of 0.94 with a highly significant Bartlett’s test of sphericity (p < 0.001) and acceptable Cronbach’s alpha coefficients indicated our constructs’ validity and internal consistency. However, it should be noted that the TPB is most utilized to predict pro-social behavioral intentions [39], with only a few studies using the framework to explain oppositional behaviors [40]. Further research applications using the TPB to test oppositional behavioral intentions are warranted to validate its use in similar contexts. Another commonly cited limitation of behavioral intention scales, such as the TPB, is that it evaluates psychological antecedents of behavioral intention, which falls short of predicting actual behavior. An area of future research is to examine how behavioral intention to oppose tourism translates to actual oppositional behavior.

Another notable limitation is that our research design relied on convenience sampling to distribute the questionnaire, as respondents were recruited solely from either Facebook groups or Reddit pages. Though we attempted to collect responses from individuals representing demographically diverse populations (i.e., posting the questionnaire to approximately 15 community groups per county), we acknowledge this approach excluded residents who are not active on social media. Future researchers should incorporate a more random sampling strategy and employ other survey distribution methods to ensure community members’ perceptions are fully captured.

It is also important to note the inherently dynamic nature of the COVID-19 pandemic, in which the general public’s understanding is ever-evolving; therefore, it is a reasonable assumption that individuals’ responses to the constructs within our model could also evolve. Given our data are cross-sectional, the scope of this research is limited to individuals’ responses that are also dynamic. Investigators could strengthen similar research models by incorporating data collection longitudinally across time.

Lastly, some studies (c.f., [41]) found residents’ age and education to impact support for tourism significantly. Therefore, the theoretical framework of this study could be further expanded by examining the potential moderating effect of these two variables on model constructs. In addition, given the social divisiveness surrounding COVID-19, it would be reasonable to include other demographic moderators such as gender, job sector, and political affiliation.

Despite the limitations outlined above, we are confident this research contributes significantly to the emerging body of work that focuses on residents’ perspectives as we move toward the post-COVID-19 era of tourism. Additionally, as the industry moves toward economic recovery, we implore D.M.O.s, business owners, and other community leaders to consider residents’ perspectives when implementing future tourism policies and regulations.