There are very few studies focussed on the hotel industry, especially on how to cope with anepidemic crisis. Hotel industry stakeholders are now looking for executive strategies forcrisis management practices in response to the pandemic. There is thus, an urgent need tofill this research gap.The World Health Organisation (WHO) released the first situation report on 2019-nCoVon 21 January 2020 (WHO, 2020a), updated its assessment of the outbreak as a public healthemergency of international concern on 31 January (WHO, 2020b) and described the outbreakas a pandemic on 11 March (WHO, 2020c). The epicentre of the COVID-19 outbreak shiftedfrom China to Italy, Spain and the USA in just two months. In the initial stage of theepidemic outbreak (mid-January to the end of February), many tourist destinations in Asia,where the Chinese are the major visitors, executed crisis management practices for the hotelindustry in response to COVID-19. With the spread of the pandemic all over the world, thehotel industry’s road to recovery has becomemuch longer, and crisis management strategiesneed to be adjusted over time. However, previous studies in crisis management practicesonly focussed on a single point in time (Israeli, 2007; Israeli and Reichel, 2003; Israeli et al.,2011), and research on epidemic crises was conducted in the post-crisis stage (Lee, 2009;Wan, 2013). This research gap, thus, needs to be filled by having a continuing study in theduring-crisis stage to answer the research question of what crisis management practicesshould be undertaken at different times during an epidemic outbreak.This study attempts to compare how the hotel industry responded to the initial (earlyFebruary) and pandemic (early April) stages of the COVID-19 crisis. This study usesimportance-usage-performance analysis to classify six categories of practices extended fromIsraeli and Reichel’s (2003) conceptual framework into four crisis management strategies.Although researchers have studied several epidemic crises before, such as severe acuterespiratory syndrome (SARS) in 2003 (Johnson Tew et al., 2008; Zeng et al., 2005), mostepidemic crises were short. When an epidemic crisis occurs, hotel industry stakeholdersexpect the crisis will be over shortly and determine crisis management practices based onthis assumption. In this case, however, the epidemic crisis has become worse and lasted for alonger period. Besides, the right crisis management practices are time-dependent, as nosingle solution established by previous studies could be adopted for all situations. Bycomparing the changes of the initial stage with the pandemic stage, this continuing studycontributes to our knowledge of contingencies in crisis management for the hotel industrythat can fit the evolution of the crisis. It also demonstrates the processes of importance-usage-performance analysis for researchers to undertake further studies in tourism crisismanagement. Recommendations are provided for governments and hotel industrystakeholders to guide their responses to the crisis.2. Literature review2.1 Tourism crisis management in the hotel industryCrisis management is a popular topic in the tourism and hospitality sectors. Many previousstudies have emphasised the important linkage between crisis and tourism. For example, DeSausmarez (2007) argued that the damage caused by a crisis threatens not only the nationaleconomy but also the livelihoods of many tourist destinations. Therefore, it is worthmonitoring crisis management strategies to minimise the negative effects. Some crises areman-made, such as economic crisis and terrorism. Due to the high uncertainty of theeconomic environment, economic crises discourage travel (U and So, 2020), and thus, adecrease in tourism demand and the loss of qualified employees (Okumus et al., 2005). Theimpact of terrorism can, from the tourists’ view, be broader and different levels of terrorismthreats influence their selection in travel packages (Walters et al., 2019). Travel demand was,IJCHM
thus, significantly reduced for low-income countries, but not so much for high-incomecountries (Buigut et al., 2017). Some crises are natural disasters, such as epidemics. Becausesafety is the priority when tourists plan their tours, epidemic crises can reduce touristarrivals (U and So, 2020). As the demand for hotel rooms decreases (Song et al., 2011), hotelroom rates and occupancy percentage decrease accordingly (Kim et al., 2005). The impact ofan epidemic crisis can severely affect the tourism industry (Tsao and Ni, 2016) becausenormal travel patterns may take over a year to become re-established (Pine and McKercher,2004). Whatever man-made or natural crises, other than the impacts from the crisesthemselves, the responses to the crisis can also affect a destination’s image and reputation ofattractiveness, and thus, their survival and prosperity (Santana, 2004).In recent studies of crisis management in the hotel industry, researchers have examinedhow to develop a successful crisis management model. For example, Kovaltchuk et al. (2016)examined these concepts and attempted to develop the crisis management model in theRussian hotel industry. Barbe and Pennington-Gray (2018) provided insight into ways ofusing social media for hoteliers to make crisis communication during a crisis. Abo-Muradand Abdullah (2019) carried out interviews to explore the impact of turnover culture onMalaysian hotels’ crisis management. Mikulic et al. (2018) applied the principles ofintegrated risk management to make an exploratory assessment of salient risks for theCroatian tourism industry. Although these studies are valuable for the hotel industry in thepreparation for crises, such studies considered crisis management broadly. Therefore, whenone certain type of crisis such as an epidemic, happens, there is a lack of detailed crisismanagement practices for reference.Previous studies on epidemic crisis management have pointed out the severe impact onthe tourism industry and argued that more research should be conducted to understand thedifferent aspects of an epidemic crisis. In studies of SARS, Zeng et al. (2005) viewed it as ashort-term epidemic and examined the possibilities of new innovations after tourismrecovery. McKercher and Chon (2004) argued that the government should take care of over-reaction and lack of coordination that influence long-term tourism development. In studiesof Middle East respiratory syndrome (MERS), Shi and Li (2017) modelled tourist arrivalsfrom China to South Korea and found that the impact of the MERS outbreak was significantin terms of total arrivals, but was insignificant for business travel. It is obvious that theepidemic crisis is an important topic for the hotel industry, but most of the studies wereconducted after, rather than during, an epidemic crisis (Lee, 2009; Wan, 2013). Though someof the studies were conducted to examine the impact of COVID-19 (Baum and Hai, 2020;Baum et al., 2020), there is still a lack of research examining the effective crisis managementpractices during an epidemic crisis in the hotel industry. Given the seriousness of COVID-19pandemic, a timely study examining how hotels perceive and react to this crisis can provideinsight for future epidemic crises.2.2 Crisis management practicesIn the context of crisis management practices, Israeli and Reichel (2003) were the first toexamine the connection between the usage and importance of crisis management practicesin the Israeli hospitality industry. They interviewed 13 managers and, combined with astudy of the literature, constructed four categories of practices. They referred to Mansfeld(1999) and developed the “marketing” category. They divided cost cuttings into “humanresources” and “maintenance” and took “governmental assistance” as external support.Their framework consists of 21 items. Israeli and his co-authors extended the frameworkwith minor revisions to the restaurant industry (Israeli, 2007), Indian luxury hotels (Israeliet al., 2011) and travel agency sectors (Perl and Israeli, 2011). Their studies concluded thatCrisismanagementpractices
stakeholders were reactive, andmanagers strongly looked for support from the government.However, after experiencing many crises, we still do not know whether stakeholders couldlearn from the crises and change their preferences.To better understand crisis management practices, we should recognise that a crisisconsists of multiple stages, such as pre, during and post (Ritchie, 2004). The pre-disasterstage is before a triggering event, and research in the pre-disaster phase aims to develop astate of readiness (Carmeli and Schaubroeck, 2008). Research in the post-disaster phase aimsto critically examine the lessons learned and facilitate the management of future incidents;however, the research aim in the during-crisis stage is different, it is to develop emergencymanagement and contingency management (Sawalha et al., 2013). Counting the number ofstudies of crisis management practices in different stages, most of the studies wereconducted after the crisis. This study has a good value for tourism crisis research because itis conducted during the COVID-19 crisis.2.3 Importance-performance analysisThe concept of importance-performance analysis (IPA) is originated from Martilla andJames (1977) and has been widely used for tourism research for prioritising tourism serviceimprovements (Evans and Chon, 1989). The two-dimension importance-performance mapping (IP mapping) consists of four quadrants, as shown in Figure 1(a). Thex-axis is the performance and the y-axis is the importance. Service attributes of highimportance that are poorly performed fall in Quadrant I. For these, the “concentrate here” (orprioritise) strategy is applied to assign more resources to improve performance. The “keepup the good work” (or maintain) strategy is used for service attributes in Quadrant II (highimportance and high performance), which are the competitive strengths of tourism firms.The not-so-important service attributes fall in Quadrant III (low importance and lowperformance); although they are not performed well, they pose less of a threat to the tourismfirm, and thus, the “low priority” strategy is applied. However, for those well-performed butFigure 1.Traditional andadvancedimportance-performancemappingsIJCHM
not important service attributes, the “possible overkill” strategy should be used to saveresources for items in Quadrant IV. By examining the points in each quadrant, the tourismfirm can develop an action plan for improvements (Bacon, 2003).To improve its discriminative power, researchers have made different modifications tothis scheme (De Nisco et al., 2015; Rial et al., 2008). Recently, Lai and Hitchcock (2015) haveprovided a comprehensive framework for addressing the reliability and validity issues.They also introduced a “data-centred diagonal line”model, for which a diagonal line is usedto separate the “concentrate here” quadrant, as shown in Figure 1(b). Although IPA wascommonly used to measure the hotel performance by guests (Lai and Hitchcock, 2016), thereis a constraint for hotel staff taking self-evaluation in their service performance because ofthe response bias. This study also attempts to overcome this constraint to advance theframework for tourism crisis management research.3. Research method3.1 Measurement and questionnaire designA review of the social sciences citation index listed journals in the category of hospitality,leisure, sport and tourism since 2000, revealed that 4 out of 23 empirical articles wereIsraeli’s works. Israeli’s framework is a useful tool for evaluating crisis managementpractices, and this framework has dominated the research in crisis management practices.This study thus, borrows Israeli’s framework andmakes a further revision to suit the case ofCOVID-19. In-person online structured interviews were conducted with five hotel managersworking for different hotels through WhatsApp on 30 January 2020. They had 7 to 25 yearsof experience in the hotel industry and two of them worked in hotels during SARS. Theinterviews discussed themeasurable items individually. For marketing, with a specific focuson the location, they suggested that the location should be “relative safety”. For humanresources practices, according to the practices in the 2008 financial crisis, they added an item“provide voluntary early retirement or resignation plans” and modified “reducing thenumber of working days” to “reducing the number of office hours or working days” and“freezing pay rates” to “freezing or reducing pay rates”. They strongly opposed the item“firing employees to reduce labour force” because the word “firing” may create a responsebias. Therefore, this itemwas revised to “reducing the labour force by laying off employees”.For maintenance, they argued that the hotel services in the item “cost cuts by limiting hotelservices” should be related to hotel facilities. Therefore, they revised it as “cost cuts byclosing less used facilities”. For governmental assistance, because there is no municipalitytax in Macau, this item was deleted. However, they replaced “governmental assistance withcurrent expenses” to “provide subsidies to hotel employees” because the labour cost iscommonly the largest hotel expense. They also added three items that the governmentapplied during the 2008 financial crisis. These were, namely, “providing subsidies tocustomers who consume hotel rooms”, “providing subsidies to local residents who consumehotel facilities” and “holding international events to attract customers to Macau”. Due to thepolitical climate, they removed “organising a protest against the lack of governmentsupport”. Hotel cleanliness is more important in light of the COVID-19 because hotel gueststend to prioritise their health and hygiene (Jiang and Wen, 2020). In response to thepandemic, hotel managers also suggested seven practices in the category “epidemicprevention” covering disinfection and prevention procedures.Following the procedures of Lai and Hitchcock (2015) to check any misunderstandings orambiguities in the content of the 31 practices in five categories, the questionnaire wasfurther validated by two external professors. They supported the deletion of the item“organising a protest against the lack of government support” because governmentalCrisismanagementpractices
assistance should stand on what the government believes should be done. They alsosuggested that the subject of the items for governmental assistance could be changed from“industry-wide” to “government”. Based on this feedback, the revised version wastranslated into Chinese. Then, referring to Lai and Hitchcock’s (2015) guideline, the bilingualversion was tested with 10 hotel managers to evaluate its readability. They did not have anydifficulties answering the questions related to various aspects of hotel operation. A selectionof “N/A” in the demographic information was added to prevent response bias. The Chineseversion has been back translated to check the consistency with the previous translation.Table 1 shows the 31 practices for the study.The questionnaire is divided into four sections:(1) the importance of the 31 practices (scale 1–9);(2) respondents’ demographic information;(3) usage of the 31 practices (scale 1–10); and(4) perceived performance in five categories (scale 1–7).Table 1.Epidemic crisismanagementpractices for the hotelindustryCategory Practice CodeMarketingpracticesReducing prices on special offers MK1Reducing list prices of office services MK2Marketing to foreign tourists with a specific focus on the location’s distinctivefeatures and relative safetyMK3Marketing and promoting new products or services MK4Marketing to new segments MK5Marketing to domestic tourists in joint campaigns with local merchants MK6Advertising on different media channels MK7MaintenancepracticesCost cuts by postponing office maintenance MT1Cost cuts by postponing systems’maintenance MT2Cost cuts by purchasing lower-cost office supplies MT3Cost cuts by closing less used facilities MT4Extending credit or postponing scheduled payments MT5EpidemicpreventionImplement body temperature checks for customers at the entrances EP1Provide sufficient protective materials (such as masks) to employees EP2Educate employees about the knowledge of epidemic prevention EP3Remind guests of the importance of epidemic prevention EP4Implement sufficient cleaning and disinfection in public areas in hotels EP5Implement sufficient cleaning and disinfection in hotel rooms EP6Develop and implement standardised epidemic prevention procedures EP7Human resourcespracticesReducing the labour force by laying off employees HR1Reducing the labour force by unpaid vacation HR2Reducing the number of office hours or working days HR3Freezing or reducing pay rates HR4Replacing high-tenure employees with new employees HR5Increased reliance on outsourced human resources HR6Provide voluntary early retirement or resignation plans HR7GovernmentalassistanceThe government provides tax benefits GA1The government provides subsidies to hotel employees GA2The government holds international events to attract customers to Macau GA3The government provides subsidies to customers who consume hotel rooms GA4The government provides subsidies to local residents who consume hotel facilities GA5IJCHM
Because the common scale format on questionnaires is one of the potential sources ofcommon method biases (Podsakoff et al., 2003), researchers have used different point-scalesin their studies (Huang et al., 2020). This study thus, adopts different point-scales to reducecommonmethod bias and stereotypical effects.3.2 Importance-usage-performance analysisThis study is based on the framework of Israeli and Reichel (2003). IPA is appropriate foridentifying crisis management practices that should be done to cope with the COVID-19crisis because it can prioritise items for actions. However, Israeli and Reichel’s (2003)framework only measures the “usage level”. To determine “usage performance” whenmaking the IP mapping, a conversion step is suggested. Measurement of the performance ofeach category is used to obtain the indirect value of the usage performance of each crisismanagement practice. Correlation analysis is used to quantify the influence of usage level(as an independent variable) on the performance of its category (as a dependent variable) foreach practice. Because this newly developed method involves the importance of the item, itslevel of usage and the performance of its category, this method is called “importance-usage-performance” analysis. Figure 2 shows the process for this method.The correlation value between the usage of each practice and the perceived performanceof its category is the indirect measure of its performance of usage. For example, thecorrelation between the usage of MK1 and the perceived performance of marketing is theindirect measure of the performance of usage of MK1. The direct measure of importance andthe indirect measure of performance of usage for each practice are used to plot along the x-axis and the y-axis, respectively, in the IP mapping.3.3 Research settingMacau is a world gaming destination and attracted more than 30 million tourists in 2019(DSEC, 2019). The receipts of the 116 hotels and guesthouses stood at USD4.66bn in 2018(DSEC, 2018). After the initial outbreak of COVID-19 in Wuhan, China, in January 2020, theeconomy of Macau dropped greatly because over 70% of all tourists were from MainlandChina (DSEC, 2019). The Macau government, thus, took immediate action in response to theCOVID-19 crisis.In early February, after confirming 10 cases, the Macau government announced theclosure of all casinos for two weeks, as well as other entertainment facilities such ascinemas, theatres, fitness centres, bars and discos on 4 February 2020 (Macau Daily Times,2020). At the end of February, the revenue of Macau’s casinos had dropped by 88%compared with the previous year, and hotel occupancy was less than 12% (Casino Review,2020; WSJ, 2020). It was the worst ever record in Macau.Figure 2.Importance-usage-performance analysisprocessesCrisismanagementpractices
From mid-February to mid-March, there had been no new case. However, when residentsthought the government had taken an effective response to control the epidemic crisis andtourism economy could be recovered soon, imported cases began to emerge on 15 March2020. To control such imported cases, the Macau government prohibited visitors from othercountries and suspended bus links over the Hong Kong–Zhuhai–Macau Bridge on 6 April.The Macau government announced an anti-epidemic assistance fund of 10 billion MacauPatacas to help residents and businesses on 8 April (AGB, 2020). Up to 12 April, Macau hasseen an increase of 45 cases of COVID-19.This study covers two stages. The initial stage is the beginning of the outbreak ofCOVID-19. The pandemic stage is the secondwave of COVID-19 in early April.3.4 Data collectionFollowing Israeli’s studies, the targets of this study were senior hotel staff. During the crisis,online surveys were an effective way to recruit targets through existing mobile socialnetwork groups for hotels on WhatsApp and WeChat in Macau. The network groupscovered both internal firm groups and hotel association groups. All senior hotel staff inMacau received a survey invitation that explained the aims and sample targets of the study.In the initial stage, after 2 calls for the survey on mobile social network groups, only 184 setsof online questionnaires were collected through an online survey website (www.sojump.com) from 3 to 7 February 2020. Of these, 179 were valid after removing 5 problematicquestionnaires (in which most practices were marked with similar points). In the pandemicstage, 253 sets of online questionnaires were obtained from 1 to 4 April 2020. Of these, 244were valid after removing 9 problematic questionnaires. Because hotel senior staffvoluntarily responded to the call for online surveys, there was no response rate. Thebackground of the respondents in these two surveys is shown in Table A1 in the Appendix.Around 30% of the respondents were supervisorial level and 70% of respondents weremanagerial level or above. As they needed to work closely with other departments, theywere able to answer questions on various aspects of hotel operations.3.5 Follow-up in-person interviewsOn 10 to 13 June 2020, follow-up in-person online interviews were conducted with three hotelgeneral managers (who were not the same as those interviewed for the scale developmentand have not participated in the surveys) to review the practices that have been taken in thehotel industry with the results of this study. To prevent response bias, the interviewer askedwhat practices their hotels had been taking, without telling the interviewees the results ofthe study. The aim of the follow-up interviews is not to identify concepts and build theory,there is no rule for setting a particular sample size. Referring to Creswell (2007), a range ofthree to five interviews per case can prevent the bias with a single source. The duration ofin-person interviews was around an hour each. The interviews were audio-recorded, andthen the important points were converted to text. As these three general managers areresponsible for all aspects of crisis operations at their hotels, they could provide valuableinformation about their hotel crisis management practices. Therefore, follow-up interviewscan offer an explanation to the survey results in the discussion section.4. Results4.1 Descriptive analysisTable 2 shows the means and standard deviations of the importance and usage of crisismanagement practices in the two surveys. The highest values of importance and usagebelong to the epidemic prevention group in both surveys. However, the highest mean valuesIJCHM
of importance and usage in the initial survey are higher than the highest mean values in thepandemic stage (importance: 8.788 vs 8.590; usage: 9.112 vs 8.926). The lowest values ofimportance and usage belong to the human resource group in both surveys. The lowestmean values of importance and usage in the initial survey are lower than the mean values inthe pandemic stage (importance: 3.291 vs 4.086; usage: 4.117 vs 4.799). Therefore, the rangesof importance and usage in the initial stage are wider than the ranges in the pandemic stage.Table 3 shows the means and standard deviations of the perceived performance of the fivecategories in the two surveys. Respondents rated the perceived performance higher in thepandemic stage.4.2 Data quality assessmentBecause this study has revised and amended the attributes, exploratory factor analysis wasconducted using statistical package for the social sciences version 19 on the data collected inthe initial stage as recommended by Lai and Hitchcock (2015). The initial Kaiser-Meyer-Olkin (KMO) value suggested that data were adequate for factor analysis (KMO = 0.812).Table 2.Descriptive analysisInitial stage Pandemic stageImportance Usage Importance Usage(1–9) (1–10) (1–9) (1–10)Code Mean SD Mean SD Mean SD Mean SDMK1 6.101 2.415 6.302 2.572 6.906 1.867 7.086 2.001MK2 5.765 2.429 5.994 2.593 6.160 1.964 6.508 2.031MK3 7.514 1.784 6.810 2.643 7.328 1.846 7.148 2.200MK4 6.447 2.306 6.475 2.819 6.393 1.858 6.926 2.099MK5 6.665 2.203 6.726 2.641 6.619 1.771 6.943 2.009MK6 6.799 2.048 6.687 2.527 6.959 1.679 7.070 2.016MK7 6.654 2.062 6.860 2.594 6.955 1.628 7.340 1.938MT1 6.972 1.936 7.296 2.331 6.648 1.868 7.246 1.979MT2 6.274 2.215 7.145 2.434 6.074 1.953 6.943 2.089MT3 5.514 2.397 6.804 2.605 5.262 2.358 6.611 2.356MT4 7.028 1.950 7.743 2.271 6.488 1.999 7.279 2.091MT5 6.419 1.916 5.883 2.534 6.279 1.920 6.365 2.183EP1 8.575 0.847 9.112 1.572 8.377 1.183 8.840 1.747EP2 8.726 0.634 8.642 2.035 8.459 1.091 8.668 1.770EP3 8.676 0.692 8.520 2.034 8.443 1.060 8.775 1.714EP4 8.603 0.817 8.693 1.845 8.357 1.112 8.693 1.792EP5 8.788 0.540 8.989 1.608 8.590 0.969 8.926 1.553EP6 8.726 0.685 8.894 1.727 8.557 0.988 8.824 1.693EP7 8.721 0.645 8.911 1.612 8.516 1.014 8.779 1.701HR1 3.760 2.478 4.430 2.891 5.131 2.148 5.648 2.426HR2 5.402 2.388 6.101 2.915 5.783 1.848 6.184 2.370HR3 6.123 2.260 6.654 2.713 6.213 1.697 6.471 2.209HR4 3.933 2.345 4.899 3.045 5.066 2.114 5.680 2.344HR5 3.291 2.250 4.117 2.707 4.086 2.155 4.799 2.362HR6 4.173 2.442 4.922 2.726 5.184 1.845 5.611 2.156HR7 5.134 2.523 5.162 2.921 5.898 1.889 6.090 2.292GA1 7.464 1.824 6.492 2.781 7.574 1.405 7.119 2.274GA2 7.453 1.867 5.788 2.880 7.389 1.550 6.762 2.368GA3 6.687 2.516 5.966 2.920 6.561 2.114 6.143 2.626GA4 5.866 2.632 5.168 2.917 6.590 1.995 6.066 2.622GA5 6.235 2.470 5.391 2.902 6.557 1.914 6.287 2.551Crisismanagementpractices
After one cycle reduction, 30 items were retained in six categories, as shown in Table A2 inthe Appendix. The final KMO value was 0.813. The six categories were pricing (two items),marketing (five items), maintenance (four items), epidemic prevention (seven items), humanresources (seven items) and governmental assistance (five items). All factor loadings arehigher than 0.6, and the lowest one is 0.608. The convergent validity is assured (Hair et al.,2010).Confirmatory factor analysis was conducted using SmartPLS (Ringle et al., 2015) tofurther assess the quality of two sets of data, as recommended by Lai and Hitchcock (2015).The use of partial least squares-structural equation modelling is advantageous whenanalysing small sample sizes (Hair et al., 2017). As shown in Table 4, for two sets of data, thevalues for Cronbach’s alpha and the construct reliability (CR) of six factors are above therecommended reliability level of 0.7. The values of the average variance extracted (AVE) ofsix factors exceed the minimum criterion of 0.5. The square root of the AVE of each factorexceeds the correlations between it and other factors. Internal consistency, CR, convergentvalidity and discriminant validity are thus, also assured (Hair et al., 2010).Table 3.Perceivedperformance inepidemic crisismanagementInitial stage Pandemic stageCode Performance measure Mean SD Mean SDMKP I believe that my hotel will perform well inmarketing practices for this crisis5.095 1.452 5.295 1.076MTP in maintenance practices for this crisis 5.268 1.401 5.357 1.083EPP in epidemic prevention practices for thiscrisis5.542 1.403 5.668 1.075HRP in human resources practices for this crisis 5.011 1.502 5.205 1.190GAP I believe that the government will performwell in governmental assistance practicesfor this crisis5.156 1.546 5.324 1.305Table 4.Reliability, constructvalidity anddiscriminant validityPractice Cronbach’s alpha CR AVE PR MK MT EP HR GAInitial stage (early-February 2020)Pricing (PR) 0.918 0.960 0.924 0.961Marketing (MK) 0.867 0.904 0.654 0.436 0.809Maintenance (MT) 0.790 0.861 0.611 0.215 0.188 0.782Epidemic prevention (EP) 0.932 0.942 0.700 0.059 0.139 0.183 0.837Human resources (HR) 0.850 0.884 0.524 0.108 0.167 0.292 0.079 0.724Governmental assistance (GA) 0.836 0.883 0.602 0.246 0.396 0.253 0.184 0.318 0.776Pandemic stage (early-April 2020)Pricing (PR) 0.803 0.908 0.831 0.912Marketing (MK) 0.817 0.873 0.580 0.407 0.762Maintenance (MT) 0.766 0.833 0.561 0.249 0.229 0.749Epidemic prevention (EP) 0.960 0.967 0.806 0.015 0.214 0.063 0.898Human resources (HR) 0.844 0.878 0.509 0.257 0.173 0.354 0.119 0.713Governmental assistance (GA) 0.759 0.837 0.509 0.247 0.377 0.171 0.293 0.283 0.714Notes: : AVE average variance extracted; CR composite reliability; Italic front – square-root of AVEIJCHM
4.3 Importance-performance mapping in the initial stageTable 5 shows the values of the correlation between each crisis management practice andthe performance of its category. In the initial stage, the highest and lowest correlation valuesare 0.475 (between MK4 and MKP) and0.059 (between HR4 and HRP), respectively. Thesevalues are used for usage to plot IP mapping with the values of importance (from Table 2),as shown in Figure 3. The grand mean values of importance (6.602) and usage performance(0.273) are used as cross-points in the mapping. The diagonal line is the regression line withthe best fit. In total, 17, 5, 5 and 3 practices fall in Quadrant I, II, III and IV, respectively. Allitems of epidemic prevention andmaintenance are in Quadrant I.4.4 Importance-performance mapping in the pandemic stageIn the pandemic stage, the highest and lowest correlation values are 0.405 (between EP5 andEPP) and 0.092 (between HR4 and HRP), respectively. Figure 4 shows the IP mapping for thepandemic stage. The grand mean values of importance and usage performance are 6.771and 0.289, respectively. In total, 16, 3, 5 and 6 practices fall in Quadrants I, II, III and IV,respectively. All items for epidemic prevention remain in Quadrant I.4.5 Comparison between two stagesBecause the importance and level of usage were measured with different scales (1–9 scale forimportance and 1–10 scale for level of usage), the usage levels were adjusted by * 9/10 to beon the same scale as importance in calculating the difference between importance and usage.As shown in Table 6, EP1 (þ5), HR3 (þ5), MK1 (þ4) and HR6 (þ4) show the greatestpositive changes in the ranking of difference between importance and usage; MK4 (8),MT1 (5), EP3 (5) and MT4 (4) show the greatest negative changes in the ranking ofdifference between importance and usage; GA1, GA2, GA3, GA5, HR1, HR5, HR7, MK2,MK6, EP2 and MT3 show little or no changes in the ranking of difference betweenimportance and usage.Table 5.Correlation betweenthe usage and theperceivedperformanceCode MKP MTP EPP HRP GAPInitial stage (early-February 2020)MK1 0.210 MT1 0.269 EP1 0.277 HR1 0.050 GA1 0.314MK2 0.140 MT2 0.205 EP2 0.474 HR2 0.084 GA2 0.369MK3 0.421 MT3 0.082 EP3 0.405 HR3 0.066 GA3 0.350MK4 0.475 MT4 0.160 EP4 0.346 HR4 0.059 GA4 0.350MK5 0.435 EP5 0.442 HR5 0.002 GA5 0.394MK6 0.399 EP6 0.417 HR6 0.088MK7 0.416 EP7 0.472 HR7 0.145Pandemic stage (early-April 2020)MK1 0.334 MT1 0.214 EP1 0.382 HR1 0.173 GA1 0.382MK2 0.368 MT2 0.152 EP2 0.347 HR2 0.066 GA2 0.358MK3 0.352 MT3 0.183 EP3 0.329 HR3 0.119 GA3 0.260MK4 0.378 MT4 0.203 EP4 0.355 HR4 0.092 GA4 0.352MK5 0.305 EP5 0.405 HR5 0.133 GA5 0.392MK6 0.351 EP6 0.393 HR6 0.314MK7 0.370 EP7 0.397 HR7 0.201Notes: MKP – predicted performance of marketing practices; MTP – predicted performance ofmaintenance practices; EPP – predicted performance of Epidemic prevention; HRP – predicted performanceof human resources practices; GAP – predicted performance of governmental assistancesCrisismanagementpractices
Comparing the relocations of the crisis management practices into four quadrants in IPmapping, two pricing practices (MK1: I ! II and MK2: I ! IV) and one maintenancepractice (MT3: I ! III) move from Quadrant I to other quadrants. One marketing practice(MK3) and one government assistant practice (GA3) are upgraded from Quadrant II toQuadrant I. One marketing practice (MK5) is downgraded from Quadrant II to Quadrant IV.Other practices remain in the same quadrants except for HR6, which is relocated fromQuadrant III to Quadrant IV. For the distances of the crisis management practices from thediagonal line in Quadrant I, EP1, EP4, MT4, EP3 and HR3 are far from the line in the initialstage. In the pandemic stage, all items for EP, as well as HR2 and HR3, are far from thediagonal line. HR2 andMT4 show a large change.5. Discussion and conclusions5.1 ConclusionsConsidering the movements of the crisis management practices in the quadrants, most of themarketing practices are downgraded. In the initial stage, hotel managers hoped that theepidemic crisis could be resolved within a short period. Therefore, they suggested reducingthe price to increase demand. However, after reaching the pandemic stage, they knew thatmarketing tactics would be useless in increasing sales and preferred to save marketingFigure 3.The importance-performancemapping (initialstage)IJCHM
budgets for preparations after the end of the pandemic. Follow-up interviews have clarifiedthis, as managers foresaw that marketing practices would be upgraded during the endingstage of COVID-19. This occurred in the case study of Kim et al. (2005), where many hotels inKorea offered discounted rooms at the end of the SARS crisis and successfully increased theoccupancy rate to the same level as prior to the SARS outbreak. As the second survey wasconducted in the early April, hotel managers supported marketing to foreign tourists (MK3,which was upgraded to Quadrant I), as there were no Chinese tourists. Subsequently, as thepandemic became more serious in Europe and the USA, the general managers in the follow-up interviews in June did not proposemarketing tactics to foreign tourists.Rial et al. (2008) suggested the use of the distance from the diagonal line to rank thepriority of executing actions. Among human resource practices, reducing the number ofoffice hours or working days (HR3) is one of the high priority actions in both stages.However, reducing the labour force by unpaid vacation (HR2) becomes a high priority actionat the pandemic stage. Because labour costs account for roughly 50% of operating costs(Mandelbaum, 2017), executing these two practices can effectively reduce operation costsduring the crisis. Other human resource practices cannot help much in responding to thecrisis, especially replacing workers with new employees (HR5) and outsourcing (HR6). Onthe other hand, although Del Mar Alonso-Almeida and Bremser (2013) disagreed about costreduction through reducing customer added value, hotel managers in this study supportedFigure 4.The importance-performancemapping (pandemicstage)Crisismanagementpractices
limiting facility maintenance (MT1, MT2 and MT3) to save money during crisis periods.Follow-up interviews revealed that, with the longer duration of the pandemic, hotels arecutting labour and maintenance costs more seriously. Although the survey at the pandemicstage did not support providing voluntary early retirement or resignation plans (HR7), hotelgeneral managers foresaw that there would be a wave of resignations in the short-term.Based on the changes in the ranking of differences between performance and usage (asshown in Table 6), although some practices in different categories show a great change,some governmental assistance (GA1, GA2, GA3 and GA5) and human resource practices(HR1, HR5 and HR7) maintained similar rankings. In the IP mappings, six out of sevenhuman resource practices and four out of five governmental assistance practices remain inthe same quadrant. This implies that in either the initial stage or pandemic stage, thestrategies for implementing these two categories do not need to be adjusted. Follow-upinterviews explained that these actions have become common because of the 2008 financialcrisis. As governmental assistance does not cost hotel any money, hotel managers agreedthat it needed to be executed continuously. However, the locations shift closer to theTable 6.Comparison betweeninitial and pandemicstagesCode Initial stage Pandemic stageDifference in rankingInitial stage Pandemic stageRanking(a)(b) Ranking Quadrant QuadrantMK1 18 14 4 I IIMK2 21 20 1 I IVMK3 3 6 3 II IMK4 14 22 8 IV IVMK5 15 18 3 II IVMK6 9 10 1 II IIMK7 17 19 2 II IIMT1 19 24 5 I IMT2 25 28 3 I IMT3 30 30 0 I IIIMT4 23 27 4 I IEP1 20 15 5 I IEP2 8 7 1 I IEP3 7 12 5 I IEP4 10 13 3 I IEP5 13 11 2 I IEP6 11 8 3 I IEP7 12 9 3 I IHR1 26 25 1 III IIIHR2 24 21 3 I IHR3 22 17 5 I IHR4 29 26 3 III IIIHR5 28 29 1 III IIIHR6 27 23 4 III IVHR7 16 16 0 III IIIGA1 2 2 0 I IGA2 1 1 0 I IGA3 5 4 1 II IGA4 6 3 3 IV IVGA5 4 5 1 IV IVNotes: (a)Ranking is the ranking of difference between performance and usage. (b) difference betweenperformance and usage = performance – usage * 9/10IJCHM
diagonal line, indicating a reduced priority because the government had already releasedpolicies to help residents. Hotel managers, thus do not expect any further assistance fromthe government. The significance of governmental assistance may be downgraded in theending stage of the epidemic. For human resource practices, it is hard to recruit skilledworkers in Macau because of the lack of labour force. Therefore, hotel managers want tomaintain strategies to minimise the impact on staff in both stages of the crisis.5.2 Theoretical implicationsThe original design of IPA commonly uses direct measures for both importance andperformance of each tourism service to construct an IP map (Lai and Hitchcock, 2015).Because tourists are customers, there is no conflict in measuring tourism service providers’performance. However, in this case, senior hotel staff may have conflicts in directlymeasuring the performance of some practices. To eliminate response bias, this studymeasured the level of usage of each practice and then correlated this with the perceivedperformance of its category. Matzler et al. (2003) supported the use of correlation analysis toget indirect values because it can eliminate multi-collinearity. Furthermore, hotel managersrated the importance of practices based on the industrial environment. Direct measures canthus, be used. However, as each hotel may have different usage levels, indirect measurescorrelated with the perceived performance are appropriate. This study demonstrates howimportance-usage-performance analysis works in avoiding response bias. The applicationsof importance-usage-performance analysis are not limited to tourism crisis management,and researchers are encouraged to use this method for tourism research in other contexts.Through in-person interviews, this study extends Israeli and Reichel’s (2003) conceptualframework. Apart from adding seven epidemic prevention practices, three items of humanresources are revised with four items of governmental assistance added. Finally, the numberof practices was increased from 21 to 30. The results show that all epidemic preventionpractices fall in Quadrant I in both the initial and pandemic stages and should be the toppriority. Given COVID-19 is a hot topic in tourism research, this study contributes to ameasurement scale for epidemic prevention practices that can be used for COVID-19 relatedtourism research. The study by Israeli et al. (2011) on Indian luxury hotels during or shortlyafter terrorist attacks showed that consistent with Israeli’s previous studies, managersfocussed on government support. When only four of Israeli’s categories are considered, thepriority ranking of crisis management practices, according to the distance from the diagonalline, showed the priority of governmental assistance is lower than some cost reductionpractices (HR2, HR3, MT1, MT2 and MT4). This implies that, unlike 10 years ago,stakeholders have gained knowledge from previous crises and selected proactive crisismanagement practices. Hotel operators understand that to reduce loss and recover fasterthan their competitors, they should not rely only on government actions.As noted above, the significance of certain practices varies in the two stages. Thedemand for governmental assistance, for example, declined in the pandemic stage. Althoughresearchers commonly agreed that different crisis management strategies should beconsidered in different periods of the crisis, there is a lack of studies crossing the crisisperiods to explain such changes. According to the results, the respondents’ perception of theimportance of practices changed over time (Figure 5). The follow-up interviews providedinsight to explain the changes. For practices that have rooms for executive oversight, suchas human resources, the perception of their importance increased. On the other hand, forpractices that have nothing more to be done, such as maintenance and epidemic prevention,their perceived importance decreased. This study explores changes in crisis managementCrisismanagementpractices
practices that are useful for researchers to develop the theories of contingency in crisismanagement.5.3 Practical implicationsPrevious studies in crisis management practices focussed on the government’s reactiveactions. This study changes the view of hotel managers from reactive to proactive inundertaking epidemic prevention, human resource, maintenance and marketing practices.For example, in the initial stage, respondents believed that “body temperature checks forcustomers” (EP1) should be given the highest priority. In the pandemic stage, however, theywere more concerned about educating “employees about the knowledge of epidemicprevention” (EP3). This indicates that managers changed their attitudes from reactive toproactive. Follow-up interviews reflected that hotel staff were no longer in a passive statewhen dealing with this epidemic crisis. Appropriate training should be arranged for staff intaking precautions against epidemics. The training should include outbreak communicationguidelines, standard precautions, infection control and hygiene techniques. Providingtraining helps reduce the likelihood employees and guests get the virus, while also buildingthe hotel’s reputation.In the initial stage, reducing the number of office hours or working days (HR3) was astrategy applied. Hotels should manage their operations to ensure good performance bylimiting hours to necessary work, such as security and reducing hours for less importanttasks such as laundry service. Taking voluntary no-pay leave (HR2) could also beconsidered if the outbreak continues. Follow-up interviews confirmed these twoarrangements and revealed that some colleagues were seeking jobs. Therefore, the forecastof “resignation plans” (HR7) becomes important in the later stages of the crisis. Furthermore,senior management should consider taking a voluntary 10%–30% pay cut to show theirsupport for staff.Reducing maintenance costs should be undertaken in both stages. In the initial stage,respondents considered closing less-used facilities (MT4). However, as the outbreakcontinues, hotels should cut costs by taking actions in all maintenance tasks except“purchasing lower-cost office supplies” (MT3). Hotels could postpone maintenance taskswithout reducing their service quality. For some facilities that involve safety considerations,such as elevators, hotels should keep their maintenance as normal.The governmental assistance to stimulate internal consumptions (GA4 and GA5) mightnot work during the crisis. The occupancy rate of hotels in Macau dropped to 11% in theinitial stage and the five-star hotels stood at 6.6% in the pandemic stage. Because the hotelindustry needs to continue coping with the devastating impact of COVID-19, thegovernment may consider different forms of support other than tax benefits (GA1) to thehotel industry. These actions may not immediately help hotel employees, however, becauseFigure 5.Changes ofimportance of thepracticesIJCHM
the bonus from gambling revenues is a large part of the income for employees in casinohotels. Without such bonuses, some employees cannot afford their heavy home mortgagesand childcare expenses. Given the strong fiscal reserves in Macau, the government shouldmaintain subsidies (GA2) for a period of time until the end of the COVID-19 pandemic.Follow-up interviews revealed that some hotel managers foresaw that the outbreak of theepidemic in China could be controlled in summer. Therefore, the government shouldconsider arranging small professional meetings to attract visitors to Macau (GA3). Thegovernment should also make a step-by-step plan to help hotels return to their normaloperations in the comingmonths.Although marketing practices are not emphasised in either the initial or pandemicstages, there was a trend in increasing the importance of marketing practices in the recoverystage. Hotel marketers should prepare promotional plans once the travel ban is lifted. Hotelscould even develop and sell special room packages offering discounts for future stays toincrease cash flow and speed up the post-crisis recovery.5.4 Limitations and further researchInvitations were sent to the same social groups on WhatsApp and WeChat for two onlinesurveys because it was hard to implement the face-to-face method during the epidemic.However, more hotel managers filled in the online questionnaires during the pandemicstage. The samples of the two surveys were, thus, not the same group, which might cause asampling error. Senior hotel staff voluntarily participated in the surveys, so self-selectionmight create selection bias. Some of the respondents were department managers and mightnot be familiar with the operations in other departments, causing a potential response bias.To speed up the recovery, the Macau government could release more assistance in thefuture. Further studies are recommended in the recovery stage to complete the research ofthe COVID-19 crisis cycle. This article revealed changes in strategies over time from theview of hotel managers. However, there is no guarantee that the strategic outcomes will besuperior. Future post-crisis research should be done to investigate whether these strategieslead to a higher rate of recovery inMacau than elsewhere.In the initial stage, the impact of the epidemic was mainly to tourist destinations near toChina. During the pandemic stage, it became a global disaster. To test the use of importance-usage-performance analysis, further studies are recommended in other tourist destinations.Epidemic crisis strategies for boutique hotels and resorts may vary, and future researchcould compare the differences across hotel segments.ReferencesAbo-Murad, M. and Abdullah, A.L.K. (2019), “Turnover culture and crisis management: insights fromMalaysian hotel industry”,Academy of Strategic Management Journal, Vol. 18 No. 2, pp. 1-14.AGB (2020), “Macau government reveals specifics of covid-19 economic support plan”, available at: https://agbrief.com/headline/macau-government-reveals-specifics-of-covid-19-economic-support-plan/Backer, E. and Ritchie, B.W. (2017), “VFR travel: a viable market for tourism crisis and disasterrecovery?”, International Journal of Tourism Research, Vol. 19 No. 4, pp. 400-411.Bacon, D.R. (2003), “A comparison of approaches to importance-performance analysis”, InternationalJournal of Market Research, Vol. 45 No. 1, pp. 1-15.Barbe, D. and Pennington-Gray, L. (2018), “Using situational crisis communication theory tounderstand Orlando hotels’ twitter response to three crises in the summer of 2016”, Journal ofHospitality and Tourism Insights, Vol. 1 No. 3, pp. 258-275.Crisismanagementpractices
Baum, T. and Hai, N.T.T. (2020), “Hospitality, tourism, human rights and the impact of COVID-19”,International Journal of Contemporary Hospitality Management, Vol. 32 No. 7, pp. 2397-2407.Baum, T., Mooney, S.K.K., Robinson, R.N.S. and Solnet, D. (2020), “COVID-19’s impact on thehospitality workforce – new crisis or amplification of the norm?”, International Journal ofContemporary Hospitality Management, doi: 10.1108/IJCHM-04-2020-0314.Buigut, S., Braendle, U. and Sajeewani, D. (2017), “Terrorism and travel advisory effects oninternational tourism”,Asia Pacific Journal of Tourism Research, Vol. 22 No. 10, pp. 991-1004.Carmeli, A. and Schaubroeck, J. (2008), “Organisational crisis-preparedness: the importance of learningfrom failures”, Long Range Planning, Vol. 41 No. 2, pp. 177-196.Casino Review (2020), “Macau’s gaming revenue fell 88% in February”, available at: www.casino-review.co/macau-gaming-revenue/Creswell, J.W. (2007), Qualitative Inquiry and Research Design: Choosing among Five Approaches, 2nded., Sage, Thousand Oaks, CA.De Nisco, A., Riviezzo, A. and Napolitano, M.R. (2015), “An importance-performance analysis of touristsatisfaction at destination level: evidence from Campania (Italy)”, European Journal of TourismResearch, Vol. 10, pp. 64-75.De Sausmarez, N. (2007), “Crisis management, tourism and sustainability: the role of indicators”,Journal of Sustainable Tourism, Vol. 15 No. 6, pp. 700-714.Del Mar Alonso-Almeida, M. and Bremser, K. (2013), “Strategic responses of the Spanish hospitality sector tothefinancial crisis”, International Journal of HospitalityManagement, Vol. 32No. 1, pp. 141-148.DSEC (2018), “Hotels and similar establishments survey 2018”, available at: www.dsec.gov.mo/getAttachment/8a076b08-86f8-40fa-81e9-19d9748aee26/E_HOT_FR_2018_Y.aspxDSEC (2019), “Tourism, MICE and gaming visitor arrivals”, available at: www.dsec.gov.mo/en-US/Statistic?id=402Evans, M.R. and Chon, K.-S. (1989), “Formulating and evaluating tourism policy using importance-performance analysis”,Hospitality Education and Research Journal, Vol. 13 No. 3, pp. 203-213.Hair, J.F., Black, W., Babin, B. and Anderson, R. (2010), Multivariate Data Analysis, 7th ed., Prentice-Hall, Upper Saddle River, NJ.Hair, J.F., Jr, Hult, G.T.M., Ringle, C.M. and Sarstedt, M. (2017), A Primer on Partial Least SquaresStructural EquationModelling (PLS-SEM), 2nd ed., Sage, Thousand Oaks.Hilliard, T.W., Scott-Halsell, S. and Palakurthi, R. (2011), “Core crisis preparedness measuresimplemented by meeting planners”, Journal of Hospitality Marketing and Management, Vol. 20No. 6, pp. 638-660.HNN (2020), “Coronavirus effects color Hilton’s 2020 outlook”, Hotel News Now, available at: www.hotelnewsnow.com/Articles/300258/Coronavirus-effects-color-Hiltons-2020-outlook (accessed 11February 2020).Huang, G.I., Chen, Y.V. and Wong, I.A. (2020), “Hotel guests’ social commerce intention: the role ofsocial support, social capital and social identification”, International Journal of ContemporaryHospitality Management, Vol. 32 No. 2, pp. 706-729.Israeli, A.A. (2007), “Crisis-management practices in the restaurant industry”, International Journal ofHospitality Management, Vol. 26 No. 4, pp. 807-823.Israeli, A.A., Mohsin, A. and Kumar, B. (2011), “Hospitality crisis management practices: the case of Indianluxury hotels”, International Journal ofHospitalityManagement, Vol. 30No. 2, pp. 367-374.Israeli, A.A. and Reichel, A. (2003), “Hospitality crisis management practices: the Israeli case”,International Journal of Hospitality Management, Vol. 22 No. 4, pp. 353-372.Jiang, Y. and Wen, J. (2020), “Effects of COVID-19 on hotel marketing and management: a perspectivearticle”, International Journal of Contemporary Hospitality Management, Vol. 32 No. 8,pp. 2563-2573.IJCHM
Johnson Tew, P., Lu, Z., Tolomiczenko, G. and Gellatly, J. (2008), “SARS: lessons in strategic planningfor hoteliers and destination marketers”, International Journal of Contemporary HospitalityManagement, Vol. 20 No. 3, pp. 332-346.Kim, S.S., Chun, H. and Lee, H. (2005), “The effects of SARS on the Korean hotel industry and measuresto overcome the crisis: a case study of six Korean five-star hotels”, Asia Pacific Journal ofTourism Research, Vol. 10 No. 4, pp. 369-377.Kovaltchuk, A.P., Dedusenko, E.A., Blinova, E.A. and Miloradov, K.A. (2016), “Concept and proceduresof crisis management in Russian hotel enterprises”, Journal of Environmental Management andTourism, Vol. 7 No. 3, pp. 473-480.Lai, I.K.W. and Hitchcock, M. (2015), “Importance–performance analysis in tourism: a framework forresearchers”,TourismManagement, Vol. 48, pp. 242-267.Lai, I.K.W. and Hitchcock, M. (2016), “A comparison of service quality attributes for stand-alone andresort-based luxury hotels in Macau: 3-dimensional importance-performance analysis”, TourismManagement, Vol. 55, pp. 139-159.Lee, K. (2009), “How the Hong Kong Government lost the public trust in SARS: insights for governmentcommunication in a health crisis”, Public Relations Review, Vol. 35 No. 1, pp. 74-76.McKercher, B. and Chon, K. (2004), “The over-reaction to SARS and the collapse of Asian tourism”,Annals of Tourism Research, Vol. 31 No. 3, pp. 716-719.Macau Daily Times (2020), “February 4 breaking news: Macau casinos to close for two weeks to curbcoronavirus spread”, available at: https://macaudailytimes.com.mo/breaking-news-macau-casinos-to-close-for-two-weeks-to-curb-coronavirus-spread.htmlMandelbaum, R. (2017), “How hotel managers are coping with rising wages”, available at: https://www.hotelmanagement.net/operate/how-increases-compensation-labor-costs-are-leading-to-cost-savingsMansfeld, Y. (1999), “Cycles of war, terror and peace: determinants and management of crisis andrecovery of the Israeli tourism industry”, Journal of Travel Research, Vol. 38 No. 1, pp. 30-36.Martilla, J.A. and James, J.C. (1977), “Importance-performance analysis”, Journal of Marketing, Vol. 41No. 1, pp. 77-79.Matzler, K., Sauerwein, E. and Heischmidt, K. (2003), “Importance-performance analysis revisited: therole of the factor structure of customer satisfaction”,The Service Industries Journal, Vol. 23 No. 2,pp. 112-129.Mikulic, J., Sprcic, D.M., Holicek, H. and Prebežac, D. (2018), “Strategic crisis management in tourism:an application of integrated risk management principles to the Croatian tourism industry”,Journal of DestinationMarketing andManagement, Vol. 7, pp. 36-38.Novelli, M., Burgess, L.G., Jones, A., Ritchie, B.W. (2018), “‘No Ebola. . . still doomed’ – the Ebola-induced tourism crisis”,Annals of Tourism Research, Vol. 70, pp. 76-87.Okumus, F., Altinay, M. and Arasli, H. (2005), “The impact of turkey’s economic crisis of February 2001on the tourism industry in Northern Cyprus”,TourismManagement, Vol. 26 No. 1, pp. 95-104.Perl, Y. and Israeli, A.A. (2011), “Crisis management in the travel agency sector: a case study”, Journalof VacationMarketing, Vol. 17 No. 2, pp. 115-125.Pine, R. and McKercher, B. (2004), “The impact of SARS on Hong Kong’s tourism industry”,International Journal of Contemporary Hospitality Management, Vol. 16 No. 2, pp. 139-143.Podsakoff, P.M., MacKenzie, S.B., Lee, J.-Y. and Podsakoff, N.P. (2003), “Common method biases inbehavioral research: a critical review of the literature and recommended remedies”, Journal ofApplied Psychology, Vol. 88 No. 5, pp. 879-903.Rial, A., Rial, J., Varela, J. and Real, E. (2008), “An application of importance-performance analysis (IPA)to the management of sport centres”,Managing Leisure, Vol. 13 Nos 3/4, pp. 179-188.Ringle, C.M. Wende, S. and Becker, J.M. (2015), “SmartPLS 3”, SmartPLS GmbH, Boenningstedt,available at: www.smartpls.comCrisismanagementpractices
Ritchie, B.W. (2004), “Chaos, crises and disasters: a strategic approach to crisis management in thetourism industry”,TourismManagement, Vol. 25 No. 6, pp. 669-683.Ritchie, B.W. and Jiang, Y. (2019), “A review of research on tourism risk, crisis and disastermanagement: launching the annals of tourism research curated collection on tourism risk, crisisand disaster management”,Annals of Tourism Research, Vol. 79, pp. 1-15.Santana, G. (2004), “Crisis management and tourism”, Journal of Travel and Tourism Marketing,Vol. 15 No. 4, pp. 299-321.Sawalha, I.H.S., Jraisat, L.E. and Al-Qudah, K.A.M. (2013), “Crisis and disaster management inJordanian hotels: practices and cultural considerations”, Disaster Prevention and Management:An International Journal, Vol. 22 No. 3, pp. 210-228.Shi, W. and Li, K.X. (2017), “Impact of unexpected events on inbound tourism demand modeling:evidence of middle east respiratory syndrome outbreak in South Korea”, Asia Pacific Journal ofTourism Research, Vol. 22 No. 3, pp. 344-356.Song, H., Lin, S., Witt, S.F. and Zhang, X. (2011), “Impact of financial/economic crisis on demand forhotel rooms in Hong Kong”,TourismManagement, Vol. 32 No. 1, pp. 172-186.Tsao, C.Y. and Ni, C.C. (2016), “Vulnerability, resilience, and the adaptive cycle in a crisis-prone tourismcommunity”,Tourism Geographies, Vol. 18 No. 1, pp. 80-105.U, S.C. and So, Y.C. (2020), “The impacts of financial and non-financial crises on tourism: evidence fromMacao and Hong Kong”,TourismManagement Perspectives, Vol. 33, p. 100628.Walters, G., Wallin, A. and Hartley, N. (2019), “The threat of terrorism and tourist choice behavior”,Journal of Travel Research, Vol. 58 No. 3, pp. 370-382.Wan, Y.K.P. (2013), “A comparison of the governance of tourism planning in the two specialadministrative regions (SARs) of China–Hong Kong and Macao”, Tourism Management,Vol. 36, pp. 164-177.WHO (2020a), “Novel coronavirus (2019-nCoV) situation report – 1”, available at: www.who.int/docs/default-source/coronaviruse/situation-reports/20200121-sitrep-1-2019-ncov.pdf?sfvrsn=20a99c10_4WHO (2020b), “Novel coronavirus (2019-nCoV) situation report – 11”, available at: www.who.int/docs/default-source/coronaviruse/situation-reports/20200131-sitrep-11-ncov.pdf?sfvrsn=de7c0f7_4WHO (2020c), “WHO characterizes COVID-19 as a pandemic”, available at: www.who.int/emergencies/diseases/novel-coronavirus-2019/events-as-they-happenWSJ (2020), “Out of luck: coronavirus shuts casinos in the world’s biggest gambling hub”, available at: www.wsj.com/articles/deadly-coronavirus-forces-chinas-gambling-hub-to-shut-its-casinos-11580816447Zeng, B., Carter, R.W. and De Lacy, T. (2005), “Short-term perturbations and tourism effects: the case ofSARS in China”, Current Issues in Tourism, Vol. 8 No. 4, pp. 306-322.Zhao, Y. and Xu, H. (2020), “Chinese public attention to COVID-19 epidemic: based on social media”,medRxiv, available at: https://doi.org/10.1101/2020.03.18.20038026IJCHM
