The Moderating Effects of Resident Characteristics on Perceived Gaming Impacts and Gaming Industry Support: The Case of Macao

This article was downloaded by: [Shanghai Jiaotong University]On: 17 March 2014, At: 20:05Publisher: RoutledgeInforma Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House,37-41 Mortimer Street, London W1T 3JH, UK

Journal of Travel & Tourism MarketingPublication details, including instructions for authors and subscription information:http://www.tandfonline.com/loi/wttm20

The Moderating Effects of Resident Characteristicson Perceived Gaming Impacts and Gaming IndustrySupport: The Case of MacaoJoanne Jung-Eun Yoo, (Joe) Yong Zhou, Tracy (Ying) Lu & Taegoo (Terry) KimPublished online: 24 Feb 2014.

To cite this article: Joanne Jung-Eun Yoo, (Joe) Yong Zhou, Tracy (Ying) Lu & Taegoo (Terry) Kim (2014) The ModeratingEffects of Resident Characteristics on Perceived Gaming Impacts and Gaming Industry Support: The Case of Macao, Journal ofTravel & Tourism Marketing, 31:2, 229-250, DOI: 10.1080/10548408.2014.873314

To link to this article: http://dx.doi.org/10.1080/10548408.2014.873314

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) containedin the publications on our platform. However, Taylor & Francis, our agents, and our licensors make norepresentations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of theContent. Any opinions and views expressed in this publication are the opinions and views of the authors, andare not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon andshould be independently verified with primary sources of information. Taylor and Francis shall not be liable forany losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoeveror howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use ofthe Content.

This article may be used for research, teaching, and private study purposes. Any substantial or systematicreproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in anyform to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http://www.tandfonline.com/page/terms-and-conditions

http://www.tandfonline.com/loi/wttm20

http://www.tandfonline.com/action/showCitFormats?doi=10.1080/10548408.2014.873314

http://dx.doi.org/10.1080/10548408.2014.873314

http://www.tandfonline.com/page/terms-and-conditions

http://www.tandfonline.com/page/terms-and-conditions

THE MODERATING EFFECTS OF RESIDENTCHARACTERISTICS ON PERCEIVED GAMINGIMPACTS AND GAMING INDUSTRY SUPPORT:

THE CASE OF MACAOJoanne Jung-Eun Yoo

(Joe) Yong ZhouTracy (Ying) Lu

Taegoo (Terry) Kim

ABSTRACT. Based on a joint use of social exchange and social representation theories, this studyexamined the moderating effects of resident characteristics on the relationships between perceivedgaming impacts, gaming benefits, and industry support. The findings demonstrated that the relation-ships changed when the resident characteristics (age, education, tourism industry dependence, andcommunity attachment) were introduced to the structural model as moderators. The study provides abetter understanding of which subgroup of people within a community are more or less disposed tocertain impacts of gaming. The theoretical and practical implications were discussed.

KEYWORDS. Gaming impacts, resident perception, social exchange theory, social representationtheory, Macao

INTRODUCTION

With the rapid development of the gamingindustry worldwide, resident perceptions ofgaming impacts have attracted the attention ofmany researchers, resulting in numerous studieson the topic (e.g., Chen & Hsu, 2001; Lee &Back, 2006; Lee, Kang, Long, & Reisinger,

2010; Perdue, Long, & Kang, 1995; Roehl,1999; Vong & McCartney, 2005). Most ofthese studies classified resident perceptions ofgaming development into three main benefitand cost domains: economic, sociocultural, andenvironmental (e.g., Kang, Lee, Yoon, & Long,2008; Ko & Stewart, 2002; Lee & Back, 2003;Tosun, 2002). The theoretical foundation of this

Joanne Jung-Eun Yoo is an assistant professor from Hotel, Restaurant and Institutional Management,Lerner College of Business and Economics, University of Delaware, 14 W. Main Street, Newark, DE 19716(E-mail: [email protected]).

(Joe) Yong Zhou is an assistant professor and postgraduate program director at the Faculty ofHospitality and Tourism Management, Macao University of Science and Technology, Macao SAR,China (E-mail: [email protected]).

Tracy (Ying) Lu is an assistant professor at the School of Human Environmental Sciences, Universityof Kentucky, 102 Erikson Hall, Lexington, KY 40506 (E-mail: [email protected]).

Taegoo (Terry) Kim is an assistant professor from College of Hotel and Tourism Management, Kyung HeeUniversity, 26 Kyungheedae-ro, Dongdaemun-gu, Seoul 130–701, South Korea (E-mail: [email protected]).

Address correspondence to: (Joe) Yong Zhou, Faculty of Hospitality and Tourism Management, MacaoUniversity of Science and Technology, Macao SAR, China (E-mail: [email protected]).

Journal of Travel & Tourism Marketing, 31:229–250, 2014Copyright © Taylor & Francis Group, LLCISSN: 1054-8408 print / 1540-7306 onlineDOI: 10.1080/10548408.2014.873314

229

Dow

nloa

ded

by [

Shan

ghai

Jia

oton

g U

nive

rsity

] at

20:

05 1

7 M

arch

201

4

mailto:[email protected]





stream of research has been largely laid on theSocial Exchange Theory (SET), which explainsthe relationships between individual benefits andperceptions of economic development (Ap,1990; Gursoy & Rutherford, 2004; Lindberg &Johnson, 1997). The theory assumes that one’sperception is affected by how one perceives theexchange one makes (Gursoy, Jurowski, &Uysal, 2002; Perdue, Long, & Allen, 1990). Ifa resident perceives that the benefits are greaterthan the costs, he is inclined to get involved inthe exchange and endorse future development inthe community (Allen, Hafer, Long, & Perdue,1993). Using SET as a framework, Perdue,Long, and Kang (1999) reported a positive asso-ciation between perceived employment benefitsand quality of life. Lee and Back (2006) alsosupported SET by showing a positive relation-ship between favorable perceptions and residentsupport for casino development in South Korea.Likewise, Lee et al. (2010) confirmed the appro-priateness of SET by showing that social andeconomic impacts were important determinantsof predicting resident perceived benefits andoverall support for casino development.

However, literature also indicated that SETalone does not suffice to explain resident per-ceptions and industry support, demanding anew conceptual approach to understand a com-plete picture at both individual and communitylevels (Chhabra & Gursoy, 2007). The SocialRepresentation Theory (SRT) concerns interac-tions between individuals and their social orcultural world (Moscovici, 1984). Unlike SETthat views an individual as an isolated unit, SRTenables communication to take place amongmembers of a community by providing themwith a code for social exchange and for namingand classifying the various aspects of theirworld (Gaskell, 2001; Jaspars & Fraser, 1984).SRT is particularly appropriate when the topicof study involves multiple social perspectives tosalient issues (Billig, 1993; Zhou & Ap, 2009).

Fredline and Faulkner (2000) stated that perso-nal characteristics affect resident perceptions, withthose having similar characteristics being moresimilarly disposed to changes in a given society.A handful of studies indicated that residents withdifferent sociodemographic backgrounds showeddifferent attitudes toward gaming industry

development (Chhabra & Gursoy, 2007; Spears &Boger, 2003;Vong&McCartney, 2005). However,the inconsistent results of previous investigationscall for more research examining the moderatingeffects of resident characteristics on perceived gam-ing impacts and gaming industry support.

In consideration of the above, this study inte-grates two separate research streams of SET andSRT to understand both individual and communitylevels of resident perceived gaming impacts, gam-ing benefits, and industry support. More specifi-cally, the present study aimed to examine themoderating effects of resident sociodemographicfactors (age, education, tourism industry depen-dence, community attachment) on perceived gam-ing impacts, gaming benefits, and industry support.

Compared to previous research that focusedon gaming perceptions and industry supportfrom the perspective of the entire community,this study explored the different views of indi-viduals within the community. As Madrigal(1995) argued, the fact that people live in thesame geographical area does not mean theybelong to the same community. In any givencity or region, there may be several inter-mingled communities with different goals andopinions. It is crucial for governments, policymakers, developers, and community leaders tounderstand the goals, aspirations, and attitudesof various communities to gain wider supportfor gaming industry development. Then appro-priate communication strategies and measurescan be planned and implemented accordingly.

LITERATURE REVIEW

Resident perceived gaming impacts

Due to the rapid development of the gamingindustry, research on resident perceptions of gam-ing impacts has increased over the last fewdecades.Research by Pizam and Pokela (1985) is probablyone of the earliest studies that examined gamingimpacts from the resident perspective. Accordingto the findings, while most residents perceived thatcasino development could improve their standardof living and provide additional jobs, they fearedthat it could reduce the neighborhood quality andcompletely alter the image of their town. This study

230 JOURNAL OF TRAVEL & TOURISM MARKETING

Dow

nloa

ded

by [

Shan

ghai

Jia

oton

g U

nive

rsity

] at

20:

05 1

7 M

arch

201

4

has laid the groundwork for the succeedingresearch on gaming impacts. Some earlier studies,in particular, described the gaming industry as ananswer for economic development, more employ-ment opportunities, and increased quality of life(Jinker-Lliyd, 1996; Stokowski, 1996).

However, as the gaming industry has matured,resident attitudes changed with a change in eco-nomic conditions of the host community(Carmichael, Peppard, & Boudreau, 1996; Hsu,2000). Although residents have recognized theeconomic benefits of the gaming industry, theyalso awakened to the consequences, especiallysocial and environmental issues, associated withgaming development. Negative issues usuallybecome more prevalent after gaming has beeninitiated in the community (Nickerson, 1995). Anumber of existing studies have addressed speci-fic social and environmental issues related to thegaming industry (e.g., Back & Lee, 2005; Kanget al., 2008; Lee et al., 2010; Vong & McCartney,2005; Vong, 2009). Although factors that emergedfrom each study are slightly different, severalcommonalities exist. Social costs include crime,drugs, prostitution, corruption, and destruction offamilies, while environmental costs consist oftraffic congestion, litter, pollution and noise, anddestruction of natural environment.

The shifting attitude of local residents during atransitional time of gaming development isanothermuch discussed topic in gaming literature.Carmichael et al. (1996) investigated resident atti-tudes toward gaming development in a rural areain southeastern Connecticut, using longitudinaldata. The results indicated local residents’ grow-ing awareness of both negative effects of the rapiddevelopment and positive employment benefitsduring the three-year period of the gaming devel-opment. Hsu (2000) assessed the changes in resi-dent perceptions on impacts of riverboat casinosin Iowa and Illinois between 1993 and 1998. Theresults were consistent with previous studies thatresidents were significantly less positive about thelegalization of gaming after the five-year casinodevelopment. Likewise, Lee and Back (2006)showed that residents had significantly differentperceptions about casino impacts before and afterits development. The researchers further pointedout that positive economic impact was the mostsignificant factor in predicting resident perceived

benefits and the support level of gaming develop-ment. Vong (2009) examined the changes inMacao resident gaming attitudes and perceivedgaming impacts. The results indicated that therespondents had developed more conservativeattitudes toward the gaming development overthe five-year development period. Although thesurveyed Macao respondents agreed that theregion’s economy had improved considerablydue to the casino development, they also recog-nized that the environment had been deterioratedand the cost of living had increased significantly.

Theoretical frameworks for gamingimpacts—SET and SRT

Since SET has been predominately used as atheoretical basis for explaining resident percep-tions of tourism impacts (e.g., Ap, 1990;Gursoy & Rutherford, 2004; Jurowski &Gursoy, 2004; Lindberg & Johnson, 1997),gaming researchers adopt SET in examiningresident perceived gaming impacts (e.g.,Caneday & Zeiger, 1991; Chhabra, 2007;2009; Ham, Brown, & Jang, 2004; Kanget al., 2008; Lee & Back, 2006; Lee et al.,2010; Stitt, Nichols, & Giacopassi, 2005;Vong, 2009). Madrigal (1993) stated that thetheory provides an economic analysis of inter-actions that focus on exchange and mutual dis-pensation of rewards and costs among thoseinvolved. An individual that perceives benefitsfrom an exchange is likely to evaluate the out-comes positively, whereas one that perceivescosts is more likely to evaluate the conse-quences negatively. In the context of tourism,the theory postulates that resident attitudestoward the tourism industry and subsequentlevel of support for its development may beinfluenced by the resident’s evaluation of result-ing outcomes in the community. Andereck andVogt (2000) confirmed SET in that those resi-dents who perceived tourism positively wouldsupport specific types of tourism development.Conversely, those who do not see themselves asdirect beneficiaries of the industry developmentmay view the impacts more negatively(Andereck, Valentine, Knopf, & Vogt, 2005).

Yoo et al. 231

Dow

nloa

ded

by [

Shan

ghai

Jia

oton

g U

nive

rsity

] at

20:

05 1

7 M

arch

201

4

SET stipulates that residents would seek ben-efits in an exchange for something estimated asequal to the benefits they offer in return, such asresources provided to casino developers andgamblers. What residents offer additionally inthis exchange includes their support for appro-priate development, being hospitable, and toler-ating negative externalities created by gamingdevelopment, such as crime, drugs, and trafficcongestion. If locals perceive that the benefitsare greater than the costs, they are inclined to beinvolved in the exchange and subsequentlyendorse future gaming development in theircommunity (Allen et al., 1993). Lee and Back(2003) supported SET in that those who per-ceived they would personally benefit from thecasino development would express economicand social impacts more positively. Chhabra(2009) examined social service providers’ per-ceptions of gaming impacts and their supportfor gaming development. The results showedthat the level of resident support was drivenby perceived economic and social impacts. Amore recent study by Lee et al. (2010) alsoapplied SET to show that positive social andeconomic impacts were important determinantsof predicting resident perceived benefits, whichin turn influenced industry support.

Although a number of evidences prove thatSET is an appropriate framework for studyingresident perceptions of gaming impacts andindustry support, SET alone is not adequate toexplain reactions of residents (Chhabra &Gursoy, 2007). For example, it is often thecase that people with no apparent benefits dosupport gaming development (Caneday &Zeiger, 1991; Chhabra, 2007). Other than per-sonal benefits, personal factors such as gender,age, education, and employment status mayinfluence resident support for gaming develop-ment (Perdue et al., 1995; Spears & Boger,2003; Vong & McCartney, 2005; Vong, 2009).

Another problem with SET is that it treats allresidents in a society as a single homogeneousgroup. SRT is a framework that links individualsand groups, operating on both individual and com-munity levels (Pearce, Moscardo, & Ross, 1996).It explains how various groups of people in a givencommunity understand and respond to socialaffairs. The theory is particularly appropriate

when the topic of study involves multiple socialperspectives or accompanies conflicts because ofpotential change and uncertainty (Billig, 1993;Farr, 1993; Moscovici, 1984). For this reason, byidentifying different types and content of the socialrepresentations that different groups hold, localcommunity perceptions of any reactions to a parti-cular industry or event can be well understood(Gaskell, 2001; Jaspars & Fraser, 1984).

SRT looks beyond a measure of an indivi-dual’s position on an attitude scale. It explainshow this position relates to positions on otherscales, to relationships with values, and to theorigin of knowledge and beliefs on which theattitudes are based (Moscovici, 1988). Severaltourism studies have applied SRT to understandcommunity knowledge and experience in thetourism context (Andriotis & Vaughan, 2003;Zhang, Inbakaran, & Jackson, 2006; Zhou &Ap, 2009). From the perspective of gamingdevelopment, it is likely that several social repre-sentations may coexist within a host community.

Social representations by residentcharacteristics

Pearce et al. (1996) listed three criteria that canbe applied to identify social representations: (1) thecommonality or consensus that exists amongmem-bers of a community, (2) the connection or networkof links between the impacts and related ideas, and(3) the notion that there is a central cluster servingto portray the social representations.

Fredline and Faulkner (2000) stated that demo-graphic characteristics have a bearing on residentperceptions, with those having a similar back-ground being more similarly disposed to societyand social changes. Several sociodemographic fac-tors have been investigated for their influence onthe SET framework, such as age, education,employment status, religion, race, and ethnicity.For example, Spears and Boger (2003) found thatresident perceptions of Native American Gambling(NAG) development had a positive associationwith their gender, gaming trips, age, employmentstatus, and income level. In a study by Ham et al.(2004), education, employment status, and religionwere significantly related to the level of residentsupport for gaming development. Vong and


Dow

nloa

ded

by [

Shan

ghai

Jia

oton

g U

nive

rsity

] at

20:

05 1

7 M

arch

201

4

McCartney (2005) also revealed that resident per-ceptions of gaming impacts were different amongsubgroups of gender, age, income, educationalbackground, and marital status. Similarly,Chhabra and Gursoy (2007) showed that the raceof residents exercised a significant influence ontheir perceptions of benefits and costs and casinodevelopment support. Vong (2009) explored theinfluence of personal factors on perceived gamingimpact to find that all personal factors contributedto the differences in the perceptions among sub-groups of the community.

In contrast, Perdue et al. (1999) argued that thesociodemographic characteristics of the surveyedrespondents were not related to their perceivedimpacts of gaming when controlling for personalbenefits. Back and Lee (2005) examined the under-lying relationships among demographic variablesof casino community residents, the perceptions ofpotential benefits and costs, and the support forcasino development. While the results showedsocial and economic benefits were most significantin determining the level of support for casinodevelopment, none of the demographic variablessignificantly affected the support level. Theseinconsistent results of previous studies call formore research investigating the moderating effectsof resident sociodemographic characteristics onperceived gaming impacts and industry support.

The current study used demographic factorsof age and educational background to identifysocial representations within a host communityof Macao. Husbands (1989) argued that age andthe education level of local residents were themost important variables associated with theirperceptions of tourism effects. Hsu (1998) pro-vided further support by reporting that the edu-cation level of residents was positivelycorrelated with their gaming industry support.Teye, Sönmez, and Sirakaya (2002) also empiri-cally supported that residents with a higher levelof education held more positive attitudes towardtourism development.

Apart from the demographics, there are othervariables that may moderate resident percep-tions on gaming impacts and their industry sup-port. Numerous studies suggested thatcommunity attachment may affect resident per-ceived tourism impacts and industry support.For example, McCool and Martin (1994)

reported that residents who were stronglyattached to their community viewed tourismimpacts with more concern than those whowere less attached. Similarly, Gursoy et al.(2002) argued that the more attached peopleare to their community, the more likely theyare to perceive that the local economy needsassistance. This may be an interpretation oftheir perceptions towards the costs and benefits.If people feel that new investments are neededin their region, they are likely to evaluate thebenefits more positively by minimizing thenegative impacts. As a result, residents whoexpress a high level of attachment to their com-munity are more likely to regard tourismimpacts as both economically and socially ben-eficial (Gursoy & Rutherford, 2004).

Community attachment has been operationa-lized by duration of residency in the study areas(Andereck et al., 2005; Gursoy et al., 2002).Brougham and Butler (1981) believed that lengthof residency in a particular area as well as age,tourist pressure, and language are the major influ-ential variables on perceptions of tourism impacts.Davis, Allen, and Cosenza (1988) found that newresidents were more negative about tourismimpacts as opposed to the natives of the commu-nity. In contrast, other researchers reported thatnewcomers were more enthusiastic about tourismdevelopment than those who had lived in thecommunity for a longer time (Lankford &Howard, 1994; Ryan & Montgomery, 1994).These mixed results demand further researchinvestigating the moderating effects of residentcommunity attachment on their perceived gamingimpacts and industry support.

Several studies showed that those who areeconomically dependent on tourism-relatedbusiness were more familiar with the impactsof tourism, thus more positive about tourismdevelopment. For instance, Pizam (1978)showed that residents who were employed inthe tourism industry expressed the most positiveattitudes toward tourism, while those who werenot employed in tourism-related businessesexpressed the most negative attitudes towardthe industry. Likewise, Lankford and Howard(1994) found that residents benefiting from thetourism industry have a higher level of supportfor tourism development than those not

Yoo et al. 233

Dow

nloa

ded

by [

Shan

ghai

Jia

oton

g U

nive

rsity

] at

20:

05 1

7 M

arch

201

4

employed in the industry. Ap (1992) stated thatthe community members who have business oremployment interests in the tourism industry aregenerally more positively disposed to it, as theytrade resulting costs for benefits. On the con-trary, those who are not involved in tourismbusinesses derive no substantial direct benefits(it may even cost them) and tend to hold nega-tive perceptions (Madrigal, 1993). However,Liu and Var (1986) reported conflicting evi-dences by showing no significant differencebetween residents in tourism businesses andthose in non-tourism businesses. The relation-ship is not yet conclusive until further investi-gation into the moderating effects of tourismindustry dependence on perceived gamingimpacts and industry support.

In summary, SET alone is not sufficient toprovide a complete picture of the underlyingrelationships between perceived gamingimpacts, gaming benefits, and industry support.A handful of past studies indicated that residentswith different characteristics may have dissimilarattitudes toward gaming impacts and industrysupport (Chhabra & Gursoy, 2007; Perdueet al., 1995; Spears & Boger, 2003; Vong &McCartney, 2005). However, their divergentresults showed the impetus for this study. Basedon a joint use of SETand SRT, this study focusedon investigating the moderating effects of resi-dent characteristics on the underlying relation-ships between perceived gaming impacts,gaming benefits, and industry support. Figure 1presents the conceptual model, which comprisesthe relationships investigated in the study.

METHODOLOGY

Study site—macao SAR, China

Data of this study were collected from theresidents of Macao. Macao, as a special admin-istrative region (SAR) of China, has been well-known for its boom-and-bust gaming industryfor decades. Legalized gaming activities beganin 1847, and the licensing system was intro-duced in the late 19th century. Macao’s gamingdevelopment has started a new era since 2002,when the gaming monopoly licensing systemwas replaced by new international concession-aries invited to the competitions. Consequently,major economic and gaming indicators ofMacao from 2002 revealed an economy thathas experienced a rapid growth (Table 1).According to the World Bank (2011), Macaoranked fourth place by the Gross NationalIncome per capita (purchase power parity) in2010, overtaking its neighbor Hong KongSAR. More than 81% of the total government

TABLE 1. Major Economic and Gaming Indicators of Macao: 2002–2010

2002 2004 2006 2007 2008 2009 2010

Macao GDP (MOP *billion) 56.1 82.0 116.2 144.8 166.0 170.1 233.7Per capita GDP (in USD) 15925 22372 29088 34277 37705 39141 51214Unemployment rate (%) 6.3 4.8 3.8 3.1 3.0 3.6 2.8Gaming revenue (MOP* billion) 23.50 43.51 57.52 83.85 109.83 120.38 189.59Gaming tables 339 1092 2762 4375 4017 4770 4791Slot machines 808 2254 6546 13267 11856 14363 14050Hotel rooms 8954 10019 12954 17072 18658 20316 20998Visitor arrival (*million) 11.53 16.67 22.00 26.99 22.93 21.75 24.97

Note. USD1 = MOP8.002 as of December 31, 2010.Sources. Macao Statistics and Census Service (MSCS, 2011); Macao Government Tourist Office (MGTO, 2011).

FIGURE 1. Proposed Conceptual Model

SRT SET

PerceivedPositive Gaming

Impacts

PerceivedNegative

Gaming Impacts

PerceivedGamingBenefits

GamingIndustrySupport

Resident Characteristics(Age, Education,

Tourism Industry Dependence,Community Attachment)


Dow

nloa

ded

by [

Shan

ghai

Jia

oton

g U

nive

rsity

] at

20:

05 1

7 M

arch

201

4

revenues came from the gaming industry, andits gaming revenues surpassed that of the LasVegas Strip in 2007 (MGTO, 2011).

Today, Macao has the most extensive andlucrative gaming development in the world,particularly when considering the tiny size ofthe territory. Meanwhile, the explosive growthof gaming development in the past decade hassignificantly impacted and changed the lives of552,000 residents (by the end of 2010) inhabit-ing the 29.2 km2 territories. The rapid injectionof huge international investment and manage-ment expertise contributed a large amount ofemployment opportunities and economic bene-fits to local residents. However, rapid develop-ment comes with challenges and problems.Although most residents understand the impor-tance of gaming development to the local econ-omy, the community has been concerned withliabilities of the growth machine (Harrill, Uysal,Cardon, Vong, & Dioko, 2011). Vong (2009)reported a number of problems that the residentsexpressed, including income lagging behind ris-ing living standards and property prices,increased crime, a widening gap between thehaves and the have-nots, deteriorating environ-ment and traffic conditions, and overcrowdingin the city.

As Macao’s gaming industry reached theconsolidation stage, Macao resident attitudestoward gaming impacts might have changed(Siu, 2008). Considering its intensive gaming-based economy, Macao presents an interestingcase study on resident perceptions of gamingimpacts. Moreover, unlike previous researchbased on an aggregate perspective of residents,the current study focused on similarities anddifferences in perceived gaming impacts andindustry support among different subgroupswithin the community.

Instrument development

Measurement items for this study werederived from an extensive review of extant lit-erature relating to resident perceptions on tour-ism impacts and gaming impacts (e.g., Back &Lee, 2005; Chen & Hsu, 2001; Chhabra, 2007,2009; Hsu, 2000; Kang et al., 2008; Lee &

Back, 2003, 2006; Lee et al., 2010; Spears &Boger, 2003; Stitt et al., 2005; Vong &McCartney, 2005; Vong, 2009). Academicfaculty and Macao tourism office employeeswere invited to provide comments on whetheritems were appropriate for measuring residentperceived gaming impacts. After revising someitems based upon feedback, a total of 24 itemswere developed, and all items were formulatedas statements. Respondents were required toindicate their level of agreement with eachstatement based on a 7-point Likert scale ran-ging from 1 = very strongly disagree to 7 = verystrongly agree.

The survey questionnaire comprises twoparts. The first part is designed to measureresident perceived gaming impacts, gamingbenefits, and industry support. The second partcontains questions about residents’ sociodemo-graphic background, including gender, age,employment status, and educational back-ground. The final part of the survey asks as to(1) whether the respondent is currently workingor used to work in the tourism industry (yes orno) and (2) the length of residence in Macao(number of years). The survey instrument wasoriginally written in English, translated intoChinese by professional translators, and thentranslated back to English by the authors whoare proficient in both English and Chinese.

The survey was conducted via telephone by aprofessional survey company in Macao. Macaoresidents can be widely accessed by telephonewith the city’s high household telephone own-ership rate (312 landline subscribers per 1,000residents or a home phone subscription perhousehold in 2010) (DSEC, 2011). Using acomputer-assisted telephone interviewing(CATI) system, telephone numbers were gener-ated to ensure a random sample. Following theage criterion of the Macau Statistics and CensusService, a household member over the age of 15was invited to participate in the survey. Aresponse rate of 42.7% was reported with anaverage completing time of 13 minutes. At theend, a total of 493 useful responses were col-lected and used for data analyses.

Table 2 demonstrates that the gender ratio ofthe respondents was 41.6% (male) and 58.4%(female). In terms of age, 48.2% of the

Yoo et al. 235

Dow

nloa

ded

by [

Shan

ghai

Jia

oton

g U

nive

rsity

] at

20:

05 1

7 M

arch

201

4

respondents were younger than 40 years old,and 51.8% were 40 years old or over. Slightlyless than a quarter of the respondents (23.9%)had a college degree. More than half (57%) ofthe sampled residents have lived in Macao forless than 30 years, and only 14.4% are currentlyworking or have worked in tourism-relatedbusinesses.

Data analysis

The collected data were randomly split intotwo halves to accomplish two-factor analyses(Anderson & Gerbing, 1988). One half of thedataset (n = 246) was used to conduct explora-tory factor analysis (EFA) while the other half(n = 247) was used to perform confirmatoryfactor analysis (CFA). EFA is particularly usefulas a preliminary analysis to trim the number ofitems in measurement development prior tousing CFA to confirm the scale’s hypothesizedstructure. In the first step of data analysis, EFAwas conducted to identify the underlying dimen-sions of resident perceived gaming impacts. CFAwas then computed, and the scale was finalized.

Several researchers proposed a two-stagemodel, building process in which the measure-ment model via CFA was tested before testingthe structural model (Hair, Black, Babin,Anderson, & Tatham, 2006). Subsequently,structural relations among the latent constructswere examined to test the proposed structuralequation model (SEM) with AMOS 7.0.

After testing the main effects with the struc-tural model, the moderator effects of theselected sociodemographics were estimated.Based on a multigroup approach, the moderat-ing effects of the resident characteristics on theunderlying relationships among the constructsin the model were measured by comparing thecorresponding standardized path coefficients inthe structural model (Ahuja & Thatcher, 2005;Keil, Tan, Wei, Saarinen, Tuunainen &Wassenaar, 2000).

FINDINGS

Exploratory factor analysis

The appropriateness of EFA was first deter-mined by the Kaiser–Meyer–Olkin (KMO)measure of sampling adequacy and Bartlett’stest of sphericity. The KMO was 0.875, andBartlett’s test of sphericity was significant at alevel of p < 0.001, which justified the use ofEFA. Next, the data were subjected to EFAusing a Varimax orthogonal rotation to reducethe number of scale items. An eigenvalue of 1.0

TABLE 2. Profile of Respondents (n = 493)

Characteristics Frequency(n)

Percentage(%)

GenderMale 205 41.6Female 288 58.4AgeLess than 20 84 17.020–29 103 20.930–39 51 10.340–49 98 19.950–59 96 19.560 or older 61 12.4EducationPrimary school or below 79 16.0Junior high school 140 28.4Senior high school/Technical

school156 31.6

College/University 111 22.5Graduate school 7 1.4Employment StatusEmployed 309 62.7Student 89 18.1House wife 35 7.1Unemployed 19 3.9Retired 14 8.3Income (monthly, MOP)Below 3000 139 34.33001-6000 39 7.96001-9000 59 12.09001-12000 76 15.412001-15000 59 12More than 15000 91 18.5Years of Living in MacaoLess than 20 138 28.020-29 143 29.030-39 102 20.740-49 49 9.950-59 41 8.360 or more 20 4.1Currently Working/Worked in

the Tourism Industry?Yes 71 14.4No 422 85.6


Dow

nloa

ded

by [

Shan

ghai

Jia

oton

g U

nive

rsity

] at

20:

05 1

7 M

arch

201

4

was used for factor inclusion, while a factorloading of 0.40 was applied as a benchmark toinclude items in each factor. As indicated inTable 3, four factors were derived from the 24items of perceived gaming impacts, whichexplained 59.45% of the variance. Based onthe factor loadings and contents of the factors,the factors were labeled as Positive SocialImpacts (PSI), Negative Social Impacts (NSI),Negative Economic/Environmental Impacts(NEEI), and Positive Economic Impacts (PEI).These four factors of perceived gaming impactswere employed as exogenous latent variables inthe following SEM procedures.

Measurement model

Next, CFA was employed to validate the uni-dimensionality of the scale derived from EFA.

Consequently, 22 observed indicators associatedwith the four factors were identified (Figure 2).

The measurement model was estimatedbefore the structural model, followingAnderson and Gerbing’s (1988) two-stepapproach. The chi-square value (χ2) of the mea-surement model was 705.160 (df = 305,p < 0.001). The χ2/df value of 2.312 fell withina range of acceptable values from 2 to 5 (Hairet al., 2006). Other fit indices demonstrated thatthe measurement model fit the data reasonablywell: Q = 2.312; GFI = 0.904; RMSEA = 0.052;and CFI = 0.923.

The strength of the measurement model canbe demonstrated through measures of conver-gent and discriminant validity. Table 4 indicatesthat the standardized factor loadings were allsignificant at a level of p < 0.001. One item ofindustry support (“I will recommend other com-munities to use the gaming industry to foster

TABLE 3. Exploratory Factor Analysis of Perceived Gaming Impacts

Factors Factor loadings Eigenvalue Explained variance(%)

α

Factor 1: Positive Social Impacts 6.845 28.037 0.857Increased pride of being a member of the community 0.801More recreation opportunity 0.784Improved city appearance 0.692Improved city reputation 0.676More facilities for public services 0.671More chances to know other cultures and people 0.686More leisure/entertainment opportunity 0.642Improved service quality in Macao 0.597Factor 2: Negative Social Impacts 4.864 15.064 0.869More prostitution 0.779More usury activities 0.756More bad influence on the youth 0.734More violent crime 0.726Destruction of family/more family quarrels in the community 0.684More gambling addictions 0.673More political corruptions 0.602More gambling income leaks out of Macao 0.554Factor 3: Negative Economic/Environmental Impacts 2.035 11.317 0.764Increased cost of housing 0.742Less usage of local facilities due to increased overcrowdings 0.738Increased cost of living 0.684Increased traffic congestion 0.678More destruction of local cultural/historical environment 0.453Increased gang activities 0.431Factor 4: Positive Economic Impacts 1.208 5.032 0.659More employment opportunity 0.712More business opportunity 0.598

Note. Total explained variance (%) = 59.45.

Yoo et al. 237

Dow

nloa

ded

by [

Shan

ghai

Jia

oton

g U

nive

rsity

] at

20:

05 1

7 M

arch

201

4

economic development”) was dropped as it haslow SMC (below 0.4). The composite reliability(CR) values ranged from 0.718 (PEI) to 0.897

(PSI); all were above the recommended level of0.70 (Fornell & Larcker, 1981). The averagevariance extracted (AVE) for each construct

FIGURE 2. CFA for Resident Perceived Gaming Impacts

Negative Social Impacts

NSI1

NSI2

NSI3

NSI4

NSI5

NSI6

NSI7

NSI8

0.787

0.761

0.760

0.695

0.683

0.647

0.643

0.664

0.317

0.399

0.297

0.375

0.501

0.446

0.408

0.307

Negative Economic/

Environmental Impacts

NEEI1

NEEI2

NEEI3

NEEI4

NEEI5

0.753

0.723

0.648

0.681

0.644

0.284

0.337

0.501

0.442

0.401

Positive EconomicImpacts

PEI1

PEI2

0.759

0.651

0.402

0.317

PositiveSocial Impacts

PSI1

PSI2

PSI3

PSI4

PSI5

PSI6

PSI7

0.341

0.265

0.318

0.311

0.401

0.387

0.395

0.779

0.771

0.731

0.652

0.673

0.648

0.642

Model fit : χ2 (203) = 425.691Q = 2.097GFI = 0.928RMSEA = 0.047CFI = 0.946

Notes. PSI1 = Increased pride of being a member of the community, PSI2 = More recreation opportunity, PSI3 = Improvedcity appearance, PSI4 = Improved city reputation, PSI5 = More facilities for public services, PSI6 = More chance to knowother culture and people, PSI7 = More leisure/entertainment opportunity; NSI1 = More prostitution, NSI2 = More usuryactivities, NSI3 = More bad influence on the youth, NSI4 = More violent crime, NSI5 = Destruction of family/more familyquarrels in the community, NSI6 = More gambling addictions, NSI7 = More political corruptions, NSI8 = More gamblingincome leaks out of Macao, NEEI1 = Increased cost of housing, NEEI2 = Less usage of local facilities due to increasedovercrowdings, NEEI3 = Increased cost of living, NEEI4 = Increased traffic congestion, NEEI5 = More destruction of localcultural/historical environment, PEI1 = More employment opportunity, and PEI2 = More business opportunity.


Dow

nloa

ded

by [

Shan

ghai

Jia

oton

g U

nive

rsity

] at

20:

05 1

7 M

arch

201

4

TABLE

4.Ove

rallMea

suremen

tMod

el

Con

structs

Stand

ardize

dfactor

load

ings

SMCs

AVE

CR

(α)

Positive

Social

Impac

ts0.58

10.89

7(0.851

)PSI1.Increa

sedprideof

beingamem

berof

theco

mmun

ity0.79

60.63

4PSI2.Morerecrea

tionop

portun

ity0.77

50.60

1PSI3.Im

prov

edcity

appe

aran

ce0.72

50.52

6PSI4.Im

prov

edcity

repu

tatio

n0.65

80.43

3PSI5.Morefacilitiesforpu

blic

services

0.66

90.44

8PSI6.Morech

ance

sto

know

othe

rcu

lturesan

dpe

ople

0.66

70.44

5PSI7.Moreleisure/en

tertainm

entop

portun

ity0.63

50.40

3Neg

ativeSocial

Impac

ts0.51

80.88

6(0.875

)NSI1.Morepros

titution

0.79

20.62

7NSI2.Moreus

uryac

tivities

0.76

80.59

0NSI3.Moreba

dinflue

nceon

theyo

uth

0.75

10.56

4NSI4.Moreviolen

tcrim

e0.70

80.50

1NSI5.Des

truc

tionof

family/m

orefamily

quarrels

intheco

mmun

ity0.67

40.45

4NSI6.Morega

mblingad

dictions

0.65

20.42

5NSI7.Morepo

litical

corrup

tions

0.64

90.42

1NSI8.Morega

mblinginco

meleak

sou

tof

Mac

ao0.64

50.41

6Neg

ativeEco

nomic/Environmen

talIm

pac

ts0.52

10.83

1(0.773

)NEEI1.Increa

sedco

stof

hous

ing

0.76

80.59

0NEEI2.Le

ssus

ageof

loca

lfac

ilitie

sdu

eto

increa

sedov

ercrow

ding

s0.71

40.51

0NEEI3.Increa

sedco

stof

living

0.65

30.42

6NEEI4.Increa

sedtrafficco

nges

tion

0.70

70.50

0NEEI5.Morede

structionof

loca

lcultural/h

istoric

alen

vironm

ent

0.63

50.40

3Positive

Eco

nomic

Impac

ts0.56

90.71

8(0.670

)PEI1.Moreem

ploy

men

top

portun

ity0.74

80.56

0PEI2.Morebu

sine

ssop

portun

ity0.64

60.41

7Ben

efits

0.50

50.74

7(0.721

)BE1.

Inge

neral,Mac

aoreside

ntsha

verece

ived

afairsh

areof

thebe

nefits

from

thega

mingindu

stry

0.69

30.48

0BE2.

Mem

bers

ofmyfamily

orIha

vepe

rson

ally

bene

fitedfrom

thega

mingindu

stry

0.65

70.43

2BE3.

Ove

rall,

thede

velopm

entof

thega

mingindu

stry

hasprov

ided

bene

fits

fortheva

rious

peop

lean

dgrou

psin

the

commun

itysu

chas

thego

vernmen

t,reside

nts,

busine

ssmen

,ga

mingbu

sine

sses

,an

dso

forth

0.76

40.58

4

Support

0.56

20.73

2(0.713

)SU1.

Ove

rall,

Isu

pportthede

velopm

entof

gamingindu

stry

inMac

ao0.77

80.60

5SU2.

Iwillsa

ypo

sitivethings

abou

tthede

velopm

entof

gamingindu

stry

inMac

ao0.73

50.54

0

Notes

.Allstan

dardized

factor

load

ings

aresign

ifica

ntat

p<0.00

1leve

l.Mod

elfitmea

sures:

χ2(305

)=70

5.16

0;Q

=2.31

2;GFI=0.90

4;RMSEA=0.05

2;CFI=0.92

3.

Yoo et al. 239

Dow

nloa

ded

by [

Shan

ghai

Jia

oton

g U

nive

rsity

] at

20:

05 1

7 M

arch

201

4

was between 0.505 (benefits) and 0.581 (PSI),exceeding the acceptable value of 0.50. Hence,the measurement scales used in this study estab-lished convergent validity.

Discriminant validity was evaluated based ona series of χ2 difference tests, using measures ofeach pair of constructs. As a two-dimensionalmodel for each pair of constructs was fit, theitems representing each construct were forcedinto a single factor solution. The χ2 differencetest produced significant results for each pair ofmeasures. Imposing a single factor solution onthe two sets of items representing differentconstructs demonstrated a significant deteriora-tion of the model fit. These results provideevidence of discriminant validity (Anderson &Gerbing, 1988).

In sum, an assessment of the overall mea-surement model showed strong evidence ofreliability and validity for the operationalizationof the latent constructs.

Structural model

A structural analysis was performed to exam-ine the measurement model in predicting gam-ing benefits and industry support, using themaximum likelihood estimation method. Thegoodness of fit statistics from SEM showed

that the structural model reasonably fits thedata (Figure 3). The overall fit of the modelappeared to be acceptable with χ2 (309), Q,GFI, RMSEA, and CFI of 710.082, 2.298,0.905, 0.049, and 0.931, respectively. Theresults of the multivariate test of the structuralmodel indicated that as the model explains42.1% of the variance in the gaming benefitsconstruct and 81.9% of the variance in theindustry support construct.

As shown in Figure 3, PSI (γ = 0.496,t = 5.895, p < 0.001) and PEI (γ = 0.168,t = 1.704, p < 0.10) appeared to have significantand positive influences on gaming benefits. Therelationship between NSI (γ = −0.174,t = −2.015, p < 0.05) and benefits was statisti-cally significant and negative. However, NEEIhas no significant association with gaming ben-efits. As expected, gaming benefits had a sig-nificant and positive effect on industry support(β = 0.905, t = 14.172, p < 0.001).

Testing moderating effects of residentcharacteristics

After testing the main effects via SEM, the mod-erator effects were estimated by multiple groupanalyses. The multiple group analysis is regardedas an appropriate method where relationships

FIGURE 3. Main Effect Test

0.496(5.895)***

R2 = 0.819R2 = 0.421

Model fit: χ2 (309) = 710.082; Q = 2.298; GFI = 0.905; RMSEA = 0.049; CFI = 0.931Notes: ***p < 0.001, *p < 0.05, †p < 0.10, NS: nonsignificant.

Negative SocialImpacts

Negative Economic/Environmental

Impacts


Benefits Support


−0.174(−2.015)*

−0.113

(−0.402)NS

0.168(1.704)+

0.905(14.172)***


Dow

nloa

ded

by [

Shan

ghai

Jia

oton

g U

nive

rsity

] at

20:

05 1

7 M

arch

201

4

among latent constructs are concerned (Homburg&Giering, 2001). The moderator effects were testedby comparing the corresponding standardized pathcoefficients in the structural model (Ahuja &Thatcher, 2005; Keil et al., 2000). For statisticalcomparisons, the following procedure originallyproposed by Chin (1998) was employed:

t � Value ¼ ðPCA � PCBÞ=

spooled �ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi

1

NAþ 1

NB

� �s" #(1)

From equation (1), Spooled is expressed as:

where Spooled, pooled estimator for the variance; t,t-statistic with NA + NB-2 degrees of freedom;NA, sample size of data set for A group; NB,sample size of data set for B group; SEA, standarderror of path in structural model of A group; SEB,standard error of path in structural model of Bgroup; PCA, path coefficient in structural model ofA group; PCB, path coefficient of path in struc-tural model of B group.

Age

To test the moderator role of age, the studydivided the sample data set into two subgroups:residents who are younger than 40 years old(n = 238) and 40 or older (n = 255) groups.As shown in Figure 4 and Table 5, the analysisfound that only positive gaming impacts (PSIand PEI) had significant effects on gaming ben-efits for the younger respondents. On the otherhand, the two positive impacts and NSIappeared to have significant effects on benefitsfor the older group. The estimates for pathcoefficient from the four constructs of perceivedgaming impacts to the construct of gaming ben-efits were compared across groups. The standar-dized path coefficient from PSI to gamingbenefits for the older group (0.572) was larger

than for the younger counterpart (0.521) with asignificant t-value difference (92.017,p < 0.001). For the path from NSI to benefits,the strength was higher in the older residents(−0.206) than in the less-than-40 aged group(−0.104) with a significant t-value difference(−41.438, p < 0.001). With respect to the pathfrom PEI to benefits, the older group showed ahigher path coefficient (0.121) than the youngergroup (0.101) with a significant t-value differ-ence (2.332, p < 0.05). The path coefficientsfrom NEEI to benefits were not statisticallysignificant for both groups.

Next, the estimates for path coefficients fromthe construct of benefits to the construct of

support were compared across groups. Thepath was found to be significantly differentbetween the older (γ = 0.936, t = 13.154,p < 0.001) and the younger subgroups(γ = 0.855, t = 8.770, p < 0.001), which sup-ports the moderating role of age (t-value differ-ence = −85.421, p < 0.001). This findingimplies that the aged residents are more suppor-tive of the gaming industry than theircounterparts.

Education

For testing the moderating effect of the educa-tion level of residents on the relationships amongthe three constructs (perceived gaming impacts,gaming benefits, and industry support) in themodel, two subsamples were built: residents with-out a college degree (n = 375) and residents with acollege degree (n = 118).

In the case of education as a moderator, a com-parison of the two subgroups yielded a significantt-value difference, indicating that education playeda role of moderator. Figure 5 demonstrates that allpath coefficients of the structural model were sta-tistically significant for the group with collegedegrees, while only PSI and PEI were significantfor the less educated group. When comparing

Spooled ¼

ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi�NA � 1ð Þ= NAþNB � 2ð Þ��SE2_

Aþ�NB � 1ð Þ= NAþNB � 2ð Þ��SE2

b

n or(2)

Yoo et al. 241

Dow

nloa

ded

by [

Shan

ghai

Jia

oton

g U

nive

rsity

] at

20:

05 1

7 M

arch

201

4

coefficients of the construct of perceived gamingimpacts on the benefit construct in both groups, thestandardized path figures for the residents withcollege degrees were larger than those without acollege degree (PSI: 0.631 vs. 0436; NSI: −0.164vs. −0.133: NEEI: −0.105 vs. −0.094; PEI: 0.175vs. 0.162). As shown in Table 5, significant t-valuedifferences were found between the groups regard-ing the paths from PSI, NSI, and PEI to benefits(PSI: −48.754; NSI: 26.071; NEEI: 5.437; PEI:53.714, p < 0.001).

For the relationship between gaming benefitsand industry support in the model, education wasalso found to play a moderating role (t-valuedifference = −95.453, p < 0.001). The estimateis higher in the group with college degrees (0.935vs. 0.876), indicating that the effect of gamingbenefits on the level of industry support wasmuch stronger for the educated community.

Tourism industry dependence

In the next step of the moderator analysis, thesurvey respondents were split into two groupsby their dependence on tourism for a livelihood:no dependent (n = 422) and dependent (n = 71)

groups. The moderating role of tourism industrydependence was tested by statistically compar-ing the corresponding standardized path coeffi-cients in the structural model. As indicated inFigure 6 and Table 5, all variables of PSI, NSI,and PEI were important explanatory variablesfor both sets of samples in determining theirperceived gaming benefits. Regarding thepaths from both positive gaming impacts (PSIand PEI) to gaming benefits, the strengths werehigher in those who are economically depen-dent on the tourism industry than those who arenot (PSI: 0.853 vs. 0.527; PEI: 0.201 vs. 0.153),with significant t-value differences (PSI:−37.739; PEI: 35.431, p < 0.001). The resultsalso indicate that the coefficients from bothnegative gaming impacts (NSI and NEEI) tobenefits were significantly higher for thosewho are/were not employed in tourism-relatedbusinesses than the tourism dependent subgroup(NSI: −0.174 vs. −0.137; NEEI: −0.135 vs.−0.101). Significant t-value differences werefound at a level of p < 0.001 between NSI andbenefits (−85.305) and NEEI and benefits(14.685).

Furthermore, the results showed that the ben-efits variable had a significant effect on the

FIGURE 4. Moderator Test: Age

R2 = 0.876

PositiveEconomic

Impacts

0.521(4.085)***

R2 = 0.731R2 = 0.420Negative Social Impacts

NegativeEconomic/

EnvironmentalImpacts

Benefits Support


−0.104(−0.187)NS

−0.087(−0.905)NS

0.101(1.646)+

0.855(8.770)***

Younger than 40 (n = 238)


0.572(4.983)**

R2 = 0.444Negative Social Impacts

Negative Economic/


Benefits Support


−0.206(−2.342)*

−0.041(−0.419)NS

0.121(1.664)+

0.936

40 or older (n = 255)

Model fit:χ2 (309) = 420.858; Q = 1.362; GFI = 0.902; RMSEA = 0.037; CFI = 0.960 (Younger than 40)χ2 (309) = 451.758; Q = 1.462; GFI = 0.902; RMSEA = 0.041; CFI = 0.956 (40 or older)Notes. ***p < 0.001, *p < 0.05, †p < 0.10; NS, nonsignificant.


Dow

nloa

ded

by [

Shan

ghai

Jia

oton

g U

nive

rsity

] at

20:

05 1

7 M

arch

201

4

TABLE

5.Mod

eratingEffe

ctTes

ts

Age

Younger

than

40(n

=23

8)40

orolder

(n=25

5)t-Value

Stand

ardize

dpa

thco

efficien

t(t-Value

)Stand

ardize

dpa

thco

efficien

t(t-Value

)Differen

cetest

Pos

itive

Soc

ialImpa

cts→

Ben

efits

0.52

1(4.085

***)

0.57

2(4.983

**)

92.017

***

Neg

ativeSoc

ialImpa

cts→

Ben

efits

−0.10

4(−0.18

7NS)

−0.20

6(−2.34

2*)

−41

.438

***

Neg

ativeEco

nomic/Env

ironm

entalImpa

cts→

Ben

efits

−0.08

7(−0.90

5NS)

−0.04

1(−0.41

9NS)

NA

Pos

itive

Eco

nomic

Impa

cts→

Ben

efits

0.10

1(1.646

+)

0.12

1(1.664

+)

2.33

2*Ben

efits

→Sup

port

0.85

5(8.770

***)

0.93

6(13.15

4***)

−85

.421

***

W/o

colle

gedeg

ree(n

=37

5)With

colle

gedeg

ree(n

=11

8)t-Value

Educa

tion

Stand

ardize

dpa

thco

efficien

t(t-Value

)Stand

ardize

dpa

thco

efficien

t(t-Value

)Differen

cetest

Pos

itive

Soc

ialImpa

cts→

Ben

efits

0.43

6(3.854

***)

0.63

1(4.834

***)

−48

.754

***

Neg

ativeSoc

ialImpa

cts→

Ben

efits

−0.13

3(−1.86

4NS)

−0.16

4(−2.13

0*)

26.071

***

Neg

ativeEco

nomic/Env

ironm

entalImpa

cts→

Ben

efits

−0.09

4(1.045

NS)

−0.10

5(1.645

+)

5.43

7***

Pos

itive

Eco

nomic

Impa

cts→

Ben

efits

0.16

2(1.685

+)

0.17

5(1.702

+)

53.714

***

Ben

efits

→Sup

port

0.87

6(11.00

5***)

0.93

5(10.48

2***)

−95

.453

***

W/o

tourism

dep

enden

ce(n

=42

2)With

tourism

dep

enden

ce(n

=71

)t-Value

Tourism

Dep

enden

cyStand

ardize

dpa

thco

efficien

t(t-Value

)Stand

ardize

dpa

thco

efficien

t(t-Value

)Differen

cetest

Pos

itive

Soc

ialImpa

cts→

Ben

efit

0.52

7(6.135

***)

0.85

3(3.402

***)

−37

.739

***

Neg

ativeSoc

ialImpa

cts→

Ben

efit

−0.17

4(−2.51

2*)

−0.13

7(−0.70

5*)

−85

.305

***

Neg

ativeEco

nomic/Env

ironm

entalImpa

cts→

Ben

efit

−0.13

5(−1.65

3NS)

−0.10

1(−1.03

1NS)

14.685

***

Pos

itive

Eco

nomic

Impa

cts→

Ben

efit

0.15

3(1.647

+)

0.20

1(1.983

*)35

.431

***

Ben

efits

→Sup

port

0.85

7(13.44

3***)

0.94

8(6.301

***)

−67

.137

***

Low

communityattach

men

t(n

=26

4)Highco

mmunityattach

men

t(n

=22

9)t-Value

CommunityAttac

hmen

tStand

ardize

dpa

thco

efficien

t(t-Value

)Stand

ardize

dpa

thco

efficien

t(t-Value

)Differen

cetest

Pos

itive

Soc

ialImpa

cts→

Ben

efits

0.29

8(2.997

**)

0.63

9(4.724

***)

−31

.043

***

Neg

ativeSoc

ialImpa

cts→

Ben

efits

−0.29

8(−1.74

3+)

−0.20

4(−1.98

3*)

−39

.472

***

Neg

ativeEco

nomic/Env

ironm

entalImpa

cts→

Ben

efits

−0.01

7(−0.40

7NS)

−0.08

7(−0.67

1NS)

NA

Pos

itive

Eco

nomic

Impa

cts→

Ben

efits

0.10

4(1.968

*)0.12

8(1.905

+)

49.517

***

Ben

efits

→Sup

port

0.81

7(8.342

***)

0.93

1(10.01

3***)

−9.43

4***

Note.

***p

<0.00

1,**p<0.01

,*p

<0.05

,†p<0.10

,NS,no

nsignifica

nt;NA,no

tap

plicab

le.

Yoo et al. 243

Dow

nloa

ded

by [

Shan

ghai

Jia

oton

g U

nive

rsity

] at

20:

05 1

7 M

arch

201

4

support variable in both sets of data (no depen-dent: γ = 0.857, t = 13.443, p < 0.001; tourismdependent: γ = 0.948, t = 6.301, p < 0.001). Thepath was found to be significantly differentbetween the two groups, which supports themoderator role of residents’ tourism industrydependence (t-value difference = −67.137,p < 0.001).

Community attachment

For testing the moderating effects of commu-nity attachment, the data were divided by thelength of residency in Macao based on a mediansplit (Aiken & West, 1991): low attachmentgroup (lived in Macao for less than 27 years;n = 264) and high attachment group (lived in

FIGURE 6. Moderator Test: Tourism Industry Dependence

Model fit: χ2 (309) = 619.025; Q = 2.003; GFI = 0.910; RMSEA = 0.046; CFI = 0.935 (No Tourism Dependent)χ2 (309) = 606.208; Q = 1.962; GFI = 0.898; RMSEA = 0.049; CFI = 0.927 (Tourism Dependent)

Positive Economic Impacts

0.853(3.402)***


Negative Economic/


Benefits Support


−0.137

(−1.705)+

−0.101

(−1.031)NS

0.201(1.983)*

0.948(6.301)***

Tourism Dependent (n = 71)


0.527(6.135)***


Negative Economic/


Support


−0.174

(−2.512)*

−0.135

(−1.653)+

0.153(1.647)+

0.857(13.443)***

No Tourism Dependent (n = 422)

Benefits

Notes: ***p < 0.001, *p < 0.05, †p < 0.10; NS, nonsignificant.

FIGURE 5. Moderator Test: Education

Model fit: χ2 (309) = 574.740 ; Q = 1.860; GFI = 0.90 6; RMSEA = 0.0 46; CFI = 0.936 (Without College Degree)χ2 (309) = 474.933 ; Q = 1.537; GFI = 0. 866; RMSEA = 0.0 31; CFI = 0.975 (With College Degree)



0.436(3.854)***


NegativeEconomic/


Benefits Support


−0.133

(−1.864)N

−0.094(–1.045)NS

0.162(1.685)+

0.876(11.005)***

Without College Degree (n = 375)


0.631(4.834)***

R2= 0.874R2 = 0.460Negative Social Impacts

NegativeEconomic/


Benefits Support


–0.164(−2.130)*

−0.105(–1.645)+

0.175(1.702)+

0.935(10.482)***

With College Degree (n= 118)


Dow

nloa

ded

by [

Shan

ghai

Jia

oton

g U

nive

rsity

] at

20:

05 1

7 M

arch

201

4

Macao for 27s year or longer; n = 229). Asshown in Figure 7 and Table 5, the model wasestimated in both subsamples and compared tothe corresponding path coefficients in the struc-tural model. The results show that all path coef-ficients to benefits for the high communityattachment data set was significantly strongerthan the corresponding path value for the lowcommunity attachment data (PSI: 0.639 vs.0298; NSI: −0.204 vs. −0.103: NEEI: −0.087vs. −0.017; PEI: 0.128 vs. 0.104). A seriesof t-tests were performed to find significantt-value differences between the subgroupsregarding the paths from PSI (−31.043), NSI(−39.472) and PEI (49.517) to benefits at alevel of p < 0.001.

Finally, the estimates for path coefficientsfrom benefits to support were compared acrossgroups. The strength of the link between groupsdiffered significantly (t-value differ-ence = −9.434, p < 0.001), indicating that com-munity attachment had a moderating role in therelationship between benefits and support in theframework. Specifically, the strength was higherin the long-term residents group (γ = 0.931,t = 10.013, p < 0.001) than in the newer group(γ = 0.817, t = 8.342, p < 0.001).

CONCLUSION

Using a sample of Macao residents, thisresearch tested a moderating role of residentcharacteristics on the relationships between per-ceived gaming impacts, gaming benefits, andindustry support. In tourism literature, two the-oretical frameworks dominate in explainingresident perceptions toward tourism develop-ment: SET and SRT. These two frameworksstress different levels of perceptions: the indivi-dual level and the community level (Pearceet al., 1996). SET provides a framework toexamine the underlying relationships betweenperceived tourism impacts, tourism benefits,and industry support. While having been adominant research framework for resident atti-tudes toward tourism for decades, SET has beencriticized for not being able to yield an accuratepicture by focusing on perceptual differences atmacro level (Chhabra & Gursoy, 2007).

SRT is useful in identifying various types ofsocial representations that a particular commu-nity holds, as resident perceptions and the levelof support may not be homogenous (Gaskell,2001). As Moscovici (1973) stated, society iscomprised of different resident groups with

FIGURE 7. Moderator Test: Community Attachment

Model fit: χ2 (309) = 422.403; Q = 1.367; GFI = 0.901; RMSEA = 0.036; CFI = 0.961 (Low Community Attachment)χ2 (309) = 452.376; Q = 1.464; GFI = 0.891; RMSEA = 0.042; CFI = 0.951 (High Community Attachment)


0.298(2.997)***


Negative Economic/


Benefits Support


−0.103

(−1.743)+

−0.017

(−0.407)NS

0.104(1.968)*

0.817(8.342)***

Low Community Attachment (n = 264)


0.639(4.724)***


Negative Economic/


Benefits Support


−0.204

(−1.983)*

−0.087

(−0.671)NS

0.128(1.905)+

0.931(10.013)***

High Community Attachment (n = 229)


Yoo et al. 245

Dow

nloa

ded

by [

Shan

ghai

Jia

oton

g U

nive

rsity

] at

20:

05 1

7 M

arch

201

4

different social representations. Personal char-acteristics constitute major sources of residentsocial representations (Zhou & Ap, 2009).

This study demonstrated that the socialexchange model fitted the entire data well inexplaining Macao resident perceptions towardgaming development. Overall, Macao residentsappeared to be seriously concerned about posi-tive social and economic factors as well asnegative social and economic/environmentalaspects of gaming impacts. The perceived gam-ing impacts were found to influence residentperceptions on gaming benefits where such ben-efits also had a positive effect on their supportfor the gaming industry. These findings are lar-gely consistent with the results of previous stu-dies (Caneday & Zeiger, 1991; Chhabra, 2007;2009; Ham et al., 2004; Kang et al., 2008; Lee& Back, 2006; Stitt et al., 2005; Vong, 2009),confirming that SET is an appropriate frame-work for explaining the relationships betweenresident perceived gaming impacts, gamingbenefits, and industry support.

Furthermore, the study also demonstratedthat these relationships changed when theSRT variables were introduced as moderatorsto the structural model. There has been muchdebate about using sociodemographic variablesas indicators of differing attitudes and percep-tions of tourism for the last few decades(Pizam, 1978). With the mixed results of thepast studies, researchers were unable to reachan agreement as to whether or not residents’sociodemographic factors are significant indi-cators of certain attitudes toward tourismimpacts.

The present study suggested that personalcharacteristics of residents may indeed providea better understanding of which people are moreor less disposed to certain gaming impacts.While most members of the community werefound to be aware of the positive impacts ofgaming, others were more concerned about costfactors of gaming development. For example,younger Macao residents appeared to be fullymindful of the economic contributions of gam-ing, but they were not nearly as aware of thesocial and environmental consequences, espe-cially negative gaming impacts. Conversely, inline with previous research (Husband, 1989;

Zhou, 2010), the older respondents of thisstudy were found to be more sensitive to thesocial impacts of gaming.

The test of the moderator role of educationalso yielded an interesting finding. The resi-dents with an advanced degree appeared morelikely to notice the negative impacts of gaming,while residents without a college degree werefound to pay less attention to the possible nega-tive consequences. The less educated groupmay believe that gaming development in thecommunity could provide opportunities forthem to improve their standard of living throughjobs, so they may focus more on the positiveside of impacts on personal benefits. Harrillet al. (2011) argued that a blue-collared groupmay view the gaming industry as a means ofbecoming upwardly mobile.

As expected, people who are/were employedin the tourism industry had more positive atti-tudes toward gaming development. The findingssuggested there were differences in perceptionsof gaming’s positive impacts with regards toeconomic dependence on the industry and per-ceived benefits. The study results revealed thatthose who are greater financial beneficiaries ofgaming seem more likely to perceive a higherlevel of positive impacts. Conversely, those whowere not involved with the tourism industryappeared to be more disposed to its negativeimpacts and less supportive of the gamingindustry.

Unlike some previous studies suggesting thatcommunity attachment has a negative associa-tion with perceived tourism impacts (Lankfordand Howard 1994), this study showed that resi-dents who had lived in the community for along time were more favorable toward gamingimpacts and more supportive of the industrydevelopment. For those who had lived inMacao for a longer period, it can be understoodthat they might have become accustomed togaming as a way of life and viewed it as animportant component of the city’s economy(Vong, 2009). Although once a Portuguese col-ony for over 400 years, Chinese traditions nowdominate the Macao society. Gaming has beenprevalent among Chinese people for centuriesand perceived as a normal life activity thatinvolves the spirit of competition and risk


Dow

nloa

ded

by [

Shan

ghai

Jia

oton

g U

nive

rsity

] at

20:

05 1

7 M

arch

201

4

taking (Taormina, 2009). Lam (2008) arguedthat Chinese people’s unique form of supersti-tion involving feng shui, lucky objects, andnumbers has stimulated gaming activities. Aslegalized gaming activities have accompaniedMacao residents for more than one and a halfcenturies, residents’ daily contacts with thegaming industry are common.

The findings of this study are of both theore-tical and practical significance. First, this studyintegrated two separate research streams of SETand SRT to examine resident perceptions ongaming impacts. The joint use of the two frame-works provided an improved understanding ofthe role of residents’ characteristics in the rela-tionships between perceived gaming impacts,gaming benefits, and industry support. Severalstudies debated the potential for using socio-demographic variables as indicators of differingattitudes and perceptions of gaming impactswith a great deal of disagreement. Therefore,researchers can hardly state with confidencethat residents’ personal characteristics are sig-nificant indicators of distinct attitudes towardgaming impacts and industry support. Thestudy findings suggest that residents’ industrysupport may not only be influenced by theirperceived gaming impacts and gaming benefitsbut also by residents’ social representations. Thedual faces of gaming development depend onthe interests of different communities within ahost community (Gaskell, 2001).

For practitioners, such as local governmentofficials or developers, understanding theeffects of residents’ sociodemographics couldbe of great use, as they could then hypothesizelikely reactions of certain groups of the com-munity to future planning actions. For example,after learning that more educated and non-tourism industry related residents were moreskeptical about social impacts of gaming, policymakers and planners could invite these groupsto listen to their concerns and present themwith possible sociocultural benefits associatedwith gaming development, such as improve-ment of the community welfare and higherquality of life.

In addition, such moderator analyses mayhelp practitioners understand what is acceptableand unacceptable to certain subgroups of the

community and estimate appropriate measuresto increase the level of support from each groupof the community. It appears that governmentofficials and developers should apply differentpublic relation strategies to different groups ofresidents according to their socialrepresentations.

The study findings also suggested that someresidents appeared to recognize certain positiveimpacts but most were equally aware of nega-tive impacts. Community leaders and plannersmay consider conducting an educational pro-gram to inform residents about benefits of gam-ing development to gain their endorsement.Keeping residents informed about the positiveimpacts may sharpen their awareness of benefitsand increase the level of support. As negativeconsequences of gaming development are inevi-table, public information programs would beuseful to help the public, especially the oneswith negative perceptions, understand the fullranges of the impacts and abate their concerns.Keeping residents informed about the positiveoutcomes may sharpen their awareness of ben-efits and increase the level of support. A goodpractice is the cash-sharing scheme introducedby the Macao government in which local resi-dents get cash subsidy annually (since 2008)from the government’s fruitful gaming develop-ment return. To maximize positive socialimpacts to the community, for instance, thelocal government may create a positive educa-tional environment. Through collaborationswith community colleges and educational cen-ters in the community, a training program canbe developed for locals, which can ultimatelycreate more job opportunities.

Finally, it is hoped that the results of thisstudy offer an important caveat for policymakers and local government to ensure thatthey address the needs and concerns of variousgroups that exist within a community and inte-grate these diverse views of the community intothe actual planning process.

Despite the careful design of the study, thisstudy has certain limitations, and the resultsshould be interpreted with those limitations inmind. First, the results of this study may belimited to Macao SAR, China with its uniquehistorical and economical attributes. Future

Yoo et al. 247

Dow

nloa

ded

by [

Shan

ghai

Jia

oton

g U

nive

rsity

] at

20:

05 1

7 M

arch

201

4

research is thus encouraged to apply both the-ories of SET and SRT to investigate residentperceived gaming impacts and industry supportin other contexts. Second, the limitation of theapplied measurements should be acknowledged,as the measurement items in this study werelargely generated from the existing studies(e.g., Back & Lee, 2005; Chen & Hsu, 2001;Chhabra, 2009; Hsu, 2000; Kang et al., 2008;Lee & Back, 2006; Lee et al., 2010; Stitt et al.,2005; Vong & McCartney, 2005; Vong, 2009).While the measurements were validated, quali-tative methods such as in-depth interviews orfocus groups could help to support the conclu-sions reached with the applications of SEMtechniques. Lastly, potential issues regardingnonresponse bias should be noted, as this sam-pling issue may limit the generalizability of thestudy findings. Further, limitation in sociodemo-graphic distribution of the sample should beacknowledged. For example, 42.2% of therespondents were low-income residents withan income of less than MOP 6,000/month.Future studies with larger and more sociodemo-graphically diverse samples are required to vali-date the study results.

The future remains challenging for Macao asmore casino projects are on the way. As thesuccess of the gaming industry is largely depen-dent on support of the local community, it iscritical that resident perceptions of the industryare well understood and managed. This studycan be used as a benchmark for Macao,enabling future longitudinal and comparativeanalyses of resident perceptions and industrysupport.

REFERENCES

Ahuja, M. K., & Thatcher, J. B. (2005). Moving beyondintentions and toward the theory of trying: Effects ofwork environment and gender on post-adoption infor-mation technology use. MIS Quarterly, 29(3), 427–459.

Aiken, L. S., & West, S. G. (1991). Multiple Regression:Testing and interpreting interactions. Newbury Park,CA: Sage.

Allen, L., Hafer, H., Long, R., & Perdue, R. (1993).Rural residents’ attitudes toward recreation andtourism development. Journal of Travel Research,31(4), 27–33.

Andereck, K., Valentine, K., Knopf, R., & Vogt, C. (2005).Residents’ perceptions of community tourism impacts.Annals of Tourism Research, 32(4), 1056–1076.

Andereck, K., & Vogt, C. (2000). The relationshipbetween residents’ attitudes towards tourism and tour-ism development options. Journal of Travel Research,39(1), 27–36.

Anderson, J. C., & Gerbing, D. W. (1988). Structuralequation modeling in practice: A review and recom-mended two-step approach. Psychological Bulletin,103(3), 411–423.

Andriotis, K., & Vaughan, R. D. (2003). Urban residents’attitudes toward tourism development: The case ofCrete. Journal of Travel Research, 42(2), 172–185.

Ap, J. (1990). Residents’ perception research on the socialimpacts of tourism. Annals of Tourism Research, 17(4),610–616.

Ap, J. (1992). Residents’ perceptions on tourism impacts.Annals of Tourism Research, 19(4), 665–690.

Back, K. J., & Lee, C. K. (2005). Residents perceptions ofcasino development in Korea: The Kangwon Landcasino case. UNLV Gaming Research & ReviewJournal, 9(2), 45–53.

Billig, M. (1993). Studying the thinking society: Socialrepresentations, rhetoric, and attitudes. In G. M.Breakwell and D. Canter (Eds.), Empirical approachesto social representations (pp. 39–62). Oxford:Clarendon Press.

Brougham, J. E., & Butler, R. W. (1981). A segmentationanalysis of resident attitudes to the social, impact oftourism. Annals of Tourism Research, 7(4), 569–590.

Caneday, L., & Zeiger, J. (1991). The social, economicand environmental costs of tourism to a growing com-munity as perceived by its residents. Journal of TravelResearch, 30(2), 45–48.

Carmichael, B. A., Peppard Jr, D. M., & Boudreau, F. A.(1996). Megaresort on my doorstep: Local residentattitudes toward foxwoods casino and casino gamblingon nearby Indian Reservation Land. Journal of TravelResearch, 34(3), 9–16.

Chen, J., & Hsu, H. C. (2001). Developing and validatinga riverboat gaming impact scale. Annals of TourismResearch, 28(2), 459–476.

Chhabra, D. (2007). Exploring social exchange theorydynamics in native American casino settings. UNLVGaming Research & Review Journal, 11(2), 31–49.

Chhabra, D. (2009). Exploring casino impact perceptionsof a unique population. UNLV Gaming Research &Review Journal, 13(1), 45–54.

Chhabra, D., & Gursoy, D. (2007). Perceived impacts ofgambling: Integration of two theories. UNLV GamingResearch & Review Journal, 11(1), 27–41.

Chin, W. W. (1998). The partial least squares approach forstructural equation modeling. In G. A. Marcoulides(Ed.), Modern methods for businessresearch. Methodology for business and management


Dow

nloa

ded

by [

Shan

ghai

Jia

oton

g U

nive

rsity

] at

20:

05 1

7 M

arch

201

4

(pp. 295–336). Mahwah, NJ, USA: Lawrence ErlbaumAssociates Publishers.

Davis, D., Allen, J., & Cosenza, R. (1988). Segmentinglocal residence by their attitudes, interests and opinionstoward tourism. Journal of Travel Research, 27(2), 2–8.

DSEC (Macau Statistics & Census Service). (2011).available from http://www.dsec.gov.mo/Statistic/Demographic.aspx?lang=en-US.

Farr, R. M. (1993). Theory and method in the study ofsocial representations. In G. M. Breakwell & D. V.Canter (Eds.), Empirical approaches to social repre-sentations (pp. 15–38). Oxford: Clarendon Press.

Fornell, C., & Larcker, D. (1981). Evaluating structuralequation models with unobservable and measurementerror. Journal of Marketing Research, 18(1), 39–50.

Fredline, E., & Faulkner, B. (2000). Host CommunityReactions: A cluster analysis. Annals of TourismResearch, 27(3), 763–784.

Gaskell, G. (2001). Attitudes, social representations, andbeyond. In K. Deaux & G. Philogène (Eds.),Representations of the social (pp. 228–241). Oxford:Blackwell Publishers Ltd.

Gursoy, D., Jurowski, C., & Uysal, M. (2002). Residentattitudes: A structural modelling Approach. Annals ofTourism Research, 29(1), 79–105.

Gursoy, D., & Rutherford, D. (2004). Host attitudestoward tourism, an improved structural model. Annalsof Tourism Research, 31(3), 495–516.

Hair Jr, J. F., Black, W. C., Babin, B. J., Anderson, R. E.,& Tatham, R. L. (2006). Multivariate data analysis(6th ed.). Upper Saddle River, NJ: Prentice Hall.

Ham, S., Brown, D., & Jang, S. (2004). Proponents oropponents of casino gaming: A qualitative choicemodel approach. Journal of Hospitality & TourismResearch, 28(4), 391–407.

Harrill, R., Uysal, M., Cardon, P., Vong, F., & Dioko, L.(2011). Resident attitudes towards gaming and tourismdevelopment in Macao: Growth machine theory as acontext for identifying supporters and opponents.International Journal of Tourism Research, 13(1), 41–53.

Homburg, C., & Giering, A. (2001). Personal characteris-tics as moderators of the relationship between customersatisfaction and loyalty: An empirical analysis.Psychology and Marketing, 18(1), 43–66.

Hsu, C. H. C. (1998). Impacts of riverboat gaming oncommunity quality. Journal of Hospitality & TourismResearch, 22(4), 323–337.

Hsu, C. H. C. (2000). Residents’ support for legalized gam-ing and perceived impacts of riverboat casinos: Changesin five years. Journal of Travel Research, 38(4), 390–395.

Husbands, W. (1989). Social status and perception of tourismin zambia. Annals of Tourism Research, 16(2), 237–253.

Jaspars, J., & Fraser, C. (1984). Attitudes and social repre-sentations. In R. Farr & S. Moscovici (Eds.), SocialRepresentations (pp. 101–123). Cambridge: CambridgeUniversity Press.

Jinker-Lliyd, A. (1996). Gambling on economic develop-ment. The American City & Country, 8, 57–60.

Jurowski, C., & Gursoy, D. (2004). Distance effects onresidents’ attitudes toward tourism. Annals of TourismResearch, 31(2), 296–312.

Kang, S. K., Lee, C. K., Yoon, Y., & Long, P. T. (2008).Resident perception of the impact of limited-stakescommunity-based casino gaming in mature gamingcommunities. Tourism Management, 29(4), 681–694.

Keil, M., Tan, B. C. Y., Wei, K.-K., Saarinen, T., Tuunainen,V., & Wassenaar, A. (2000). A cross-cultural study onescalation of commitment behavior in software projects.MIS Quarterly, 24(2), 299–325.

Ko, D. W., & Stewart, W. P. (2002). A structural equationmodel of resident attitudes for tourism development.Tourism Management, 23(5), 521–530.

Lam, D. (2008). Unlocking the world of Chinese gam-bling. Global Gaming Business, 8(9), 34–35.

Lankford, S. V., & Howard, D. R. (1994). Developing atourism impact attitude scale. Annals of TourismResearch, 21(1), 121–139.

Lee, C. K., & Back, K. J. (2003). Pre- and post-casinoimpact of residents’ perception. Annals of TourismResearch, 30(4), 868–885.

Lee, C. K., & Back, K. J. (2006). Examining structuralrelationships among perceived impact, benefit, and sup-port for casino development based on 4 year longitu-dinal data. Tourism Management, 27(3), 466–480.

Lee, C. K., Kang, S. K., Long, P., & Reisinger, Y. (2010).Residents’ perceptions of casino impacts: A compara-tive study. Tourism Management, 31(2), 189–201.

Lindberg, K., & Johnson, R. (1997). Modelling residentattitudes toward tourism. Annals of Tourism Research,24(2), 402–424.

Liu, J., & Var, T. (1986). Resident attitudes toward tourismimpacts in Hawaii. Annals of Tourism Research, 13(2),193–214.

Macao Government Tourist Office (MGTO). 2011. Keyindictors. [online], available form http://industry.macautourism.gov.mo/en/page/content.php?page_id=12.Accessed on 22/10/2011.

Macao Statistics & Census Service (MSCS). 2011. Statisticsof Macao. [online] available form http://www.dsec.gov.mo/Statistic.aspx?lang=en-US Accessed on 22/10/2011

Madrigal, R. (1993). A tale of two cities. Annals ofTourism Research, 20(2), 336–353.

Madrigal, R. (1995). Residents’ receptions and the role ofgovernment. Annals of Tourism Research, 22(1), 86–102.

McCool, S., & Martin, S. (1994). Community attachmentand attitudes towards tourism development. Journal ofTravel Research, 32(3), 29–34.

Moscovici, S. (1973). Foreword. In C. Herzlich (Eds.), D.Graham, trans Health and Illness, A social psychologi-cal analysiy (pp. ix-xiv). London: Academic Press.

Moscovici, S. (1984). The Phenomenon of SocialRepresentations. In R. Farr & S. Moscovici (Eds.),

Yoo et al. 249

Dow

nloa

ded

by [

Shan

ghai

Jia

oton

g U

nive

rsity

] at

20:

05 1

7 M

arch

201

4

http://www.dsec.gov.mo/Statistic/Demographic.aspx?lang=en-US

http://www.dsec.gov.mo/Statistic/Demographic.aspx?lang=en-US

http://industry.macautourism.gov.mo/en/page/content.php?page_id=12

http://industry.macautourism.gov.mo/en/page/content.php?page_id=12

http://www.dsec.gov.mo/Statistic.aspx?lang=en-US

http://www.dsec.gov.mo/Statistic.aspx?lang=en-US

Social representations (pp. 3–69). Cambridge:Cambridge University Press.

Moscovici, S. (1988). Notes towards a description ofsocial representations. European Journal of SocialPsychology, 18(3), 211–250.

Nickerson, N. P. (1995). Tourism and gambling contentanalysis. Annals of Tourism Research, 22(1), 53–66.

Pearce, P. L., Moscardo, G., & Ross, G. (1996). Tourismcommunity relationship. Pergamon: Oxford.

Perdue, R. R., Long, P. T., & Allen, L. (1990). Residentsupport for tourism development. Annals of TourismResearch, 14(3), 420–429.

Perdue, R. R., Long, P. T., & Kang, Y. S. (1995). Residentsupport for gambling as a tourism development strat-egy. Journal of Travel Research, 34(2), 3–11.

Perdue, R. R., Long, P. T., & Kang, Y. S. (1999).Boomtown tourism and resident quality of life: Themarketing of gaming to host community residents.Journal of Business Research, 44(3), 165–177.

Pizam, A. (1978). Tourism’s impacts: The social costs tothe destination community as perceived by its resi-dents. Journal of Travel Research, 16(4), 8–12.

Pizam, A., & Pokela, J. (1985). The perceived impacts ofcasino gambling on a community. Annals of TourismResearch, 12(2), 147–165.

Roehl, W. S. (1999). Quality of life issues in a casino destina-tion:Politics, casinogambling, anddevelopment inAtlanticCity. Journal of Business Research, 44(3), 223–229.

Ryan, C., & Montgomery, D. (1994). The attitudes ofBakewell residents to tourism and numbers in commu-nity responsive tourism. Tourism Management, 15(5),358–369.

Siu, R. (2008). After the games in Macau and Singapore,what next?. Casino & Gaming International, 4,17–21.

Spears, D. L., & Boger, C. A. (2003). Residents’ percep-tions and attitudes towards Native American Gaming(NAG) in Kansas: Demographics, polices & futuredevelopment. UNLV Gaming Research & ReviewJournal, 7(2), 13–24.

Stitt, B. G., Nichols, M., & Giacopassi, D. (2005).Perception of casinos as disruptive influences in USAcommunities. International Journal of TourismResearch, 7(1), 187–200.

Stokowski, P. A. (1996). Crime patterns and gamingdevelopment in rural Colorado. Journal of TravelResearch, 24(3), 63–69.

Taormina, R. J. (2009). Social and personality correlates ofgambling attitudes and behavior among Chinese resi-dents of Macau. Journal of Social and PersonalRelationships, 26(8), 1047–1071.

Teye, V., Sönmez, S., & Sirakaya, E. (2002). Residents’attitudes toward tourism development. Annals ofTourism Research, 29(3), 668–688.

Tosun, C. (2002). Host perceptions of impacts, a compara-tive tourism study. Annals of Tourism Research, 29(1),231–253.

Vong, F. (2009). Changes in residents’ gambling attitudesand perceived impacts at the fifth anniversary ofMacao’s gaming deregulation. Journal of TravelResearch, 47(3), 388–397.

Vong, F., & McCartney, G. (2005). Mapping residentperceptions of gaming impact. Journal of TravelResearch, 44(2), 177–187.

World Bank. (2011). Gross national income per capita2010, Atlas method and PPP. [online] Retrieved fromhttp://siteresources.worldbank.org/DATASTATISTICS/Resources/GNIPC.pdf

Zhang, J., Inbakaran, R. J., & Jackson, M. S. (2006).Understanding community attitudes towards tourism andhost-guest interaction in the urban-rural border region.Tourism Geographies, 8(2), 182–204.

Zhou, Y. (2010). Resident perceptions toward the impactsof the Macao Grand Prix. Journal of Convention &Event Tourism, 11(2), 138–153.

Zhou, Y., & Ap, J. (2009). Residents’ perceptions towardsthe impacts of the Beijing 2008 Olympic Games.Journal of Travel Research, 48(1), 78–91.

SUBMITTED: April 17, 2012FIRST REVISION SUBMITTED:

September 24, 2012FINAL REVISION SUBMITTED:

January 31, 2013ACCEPTED: February 19, 2013REFEREED ANONYMOUSLY


Dow

nloa

ded

by [

Shan

ghai

Jia

oton

g U

nive

rsity

] at

20:

05 1

7 M

arch

201

4

http://siteresources.worldbank.org/DATASTATISTICS/Resources/GNIPC.pdf

http://siteresources.worldbank.org/DATASTATISTICS/Resources/GNIPC.pdf

The Moderating Effects of Resident Characteristics on Perceived Gaming Impacts and Gaming Industry Support: The Case of Macao

Documents