hosting mega events: modeling locals’ support

Annals of Tourism Research, Vol. 33, No. 3, pp. 603–623, 2006� 2006 Elsevier Ltd. All rights reserved.

Printed in Great Britain
0160-7383/$32.00
doi:10.1016/j.annals.2006.01.005www.elsevier.com/locate/atoures

HOSTING MEGA EVENTSModeling Locals’ Support

Dogan GursoyK.W. Kendall

Washington State University, USA

Abstract: This study develops and tests a structural model to assess key factors on residents’perceptions of the impacts of the 2002 Winter Olympics as a mega tourism event and howthese perceptions affect their support. The model is based on previous literature and uses datacollected during the event. Community backing for mega events is affected directly and/orindirectly by five determinants of support: the level of community concern, ecocentric values,community attachment, perceived benefits, and perceived costs. There are interactionsbetween costs and benefit factors, and support relies heavily on perceived benefits rather thancosts. Theoretical and managerial implications are discussed. Keywords: determinants of sup-port, residents’ attitudes, perceived impacts, support model, mega events. � 2006 ElsevierLtd. All rights reserved.

Resume: L’accueil des mega-evenements: modelage du soutien des habitants. Cette etudedeveloppe et met a l’essai un modele structurel pour evaluer des facteurs cle pour les percep-tions des habitants au sujet des impacts des Jeux Olympiques d’hiver 2002 comme mega-even-ement de tourisme et voir comment ces perceptions influent sur leur soutien. Le modele estbase sur la litterature anterieure et utilise des donnees cueillies pendant l’evenement. Lesoutien communautaire des mega-evenements est influe directement et/ou indirectementpar cinq determinants: niveau de preoccupation communautaire, valeurs ecocentriques,attachement communautaire, benefices percus et couts percus. Il y a des interactions entreles facteurs lies aux couts et benefices, et le soutien depend beaucoup des benefices percusplutot que des couts. On discute des implications theoriques et gestionnaires. Mots-cles:determinant de soutien, attitudes des habitants, impacts percus, modele de soutien, mega-evenement. � 2006 Elsevier Ltd. All rights reserved.

INTRODUCTION

Traditionally, mega event planning involves a predominantly politi-cal planning approach, which allows little input from local residentsapart from the initial election of political representatives (Roche1994). Veal (1994) refers to this approach as hallmark decisionmaking,where the plan to proceed with a project is made first, and attemptsare later made to justify it (Haxton 1999). Recently, a more democraticapproach to such planning has emerged as an alternative, which

Both authors are members of the School of Hospitality Business Management atWashington State University (Pullman WA 99164-4742, USA. Email <[email protected]>).Dogan Gursoy’s tourism research interests include marketing, impacts, consumer informa-tion processing, and cross-cultural studies. K. Kendall specializes in tourism, with researchinterests in consumer behavior issues, and strategic management processes in hospitality andtourism.

603

604 ATTITUDES TOWARDS MEGA EVENTS

combines both technical rationality and participatory democracy in theoverall planning process (Getz 1991; Haxton 1999; Jafari 1990). As sug-gested by Haxton (1999), the more democratic approach to megaevent planning is arguably more difficult to implement and as a resultless frequently adopted, or adopted in name only.

Successful implementation of the more democratic planning ap-proaches, such as Toronto’s bid for the 1996 Summer Olympic Gamesand Calgary Olympics, suggests that community involvement and sup-port may transform such occasions more into urban festivals likely tobecome significant urban experiences for hosts and guests (Hiller1990). While active support is likely to transform a mega sporting eventinto an urban festival, it is also possible that active opposition to host-ing it may lead to delays, legal action, and abandonment of projects.Therefore, it is important to assess the level of support/oppositionand to understand the antecedents of support/opposition by localsfor local governments, policymakers, and businesses (Haxton 1993;Hernandez, Cohen and Garcia 1996). Conceptually, this may even sug-gest some form of benchmarking or barometer approach for betterestimation of when and if there are concerns with future planned pro-ceedings. Since community involvement in planning is a relativelyrecent phenomenon, it is to be expected that research into locals’ sup-port for hosting these venues is quite limited.

In contrast, research into local residents’ support for these occasionsgenerally is abundant. Indeed, its importance has been widely recog-nized by planners and businesses that have to take the views of the hostcommunity into account for the success and sustainability of theirinvestments (Williams and Lawson 2001). This component of the‘‘environmental’’ scanning and monitoring process has become rela-tively common for strategic destination management, although withlittle formal documentation. The studies reported tend to look at tour-ism as it relates to specific communities, some of which may rely heavilyon these venues, while other research looks at communities that de-pend little on them. In the latter case, residents may be unaware ofthe magnitude of the contributions or of the negative aspects. Thefindings reported in these diverse studies suggest some inconsistenciesin the relationships, and the sophistication of the modeling has beensuspect, in fact, mostly descriptive. A more formally documentedmodeling approach may be more appropriate.

The purpose of this study is to concentrate on a mega sporting event,an obtrusive tourism venue (Webb, Campbell, Schwartz and Sechrest1971), which would be much more conspicuous to the residents’ lifespace, raising their awareness levels and probably making them morereactive. This would be an extreme case on a spectrum of ‘‘in yourface’’ or obtrusive tourism, as opposed to ‘‘small inconspicuousevents’’ that may not impinge on the locals’ life space in terms ofawareness levels and may not cause emotional, social, or cognitive reac-tions. The latter type of venues and activities may be unobtrusive in acommunity (Gursoy and Rutherford 2004). Such specification wouldallow a better understanding of the actual perceived impacts and thesupport in a magnified, obtrusive setting. Specifically, the intent of this

GURSOY AND KENDALL 605

study is to build on existing knowledge by developing a model of theeffects of key factors on residents’ perceptions of the impacts andhow these affect their attitudes toward mega events. The researchobjectives are to develop a theoretical model to examine the directand/or indirect effects of various factors on the host community’s sup-port to test and refine the proposed model using structural equationmodeling and to evaluate the strength and direction of these factorson support.

SUPPORT FOR MEGA EVENTS

Hosting mega events such as the Winter Olympics requires consider-able investment of human, financial, and physical resources from hostcommunities (Haxton 1999). A lack of coordination and cohesionwithin the host community can turn the planning process into a highlycharged political and social exercise, which may require the most con-summate skills of negotiation and consensus building from those inleadership (Haxton 1999). Underestimating the power of public de-bate and support may result in time-consuming, often bitter battlingover costs, which is likely to be fueled by media criticism, as occurredover the Montreal Olympics and Edmonton Commonwealth Games(Haxton 1999). Further evidence is found in the lay press reflectingthis trepidation associated with the Athens Summer Olympics of 2004.

Several researchers suggest that even before submitting a bid forhosting, organizers should solicit inputs from several communitygroups to prompt public debate and promote community involvement(French and Disher 1997; Lenskyj 1992). Public discussions on the ex-pected benefits and costs and widespread community involvement arelikely to result in a broad public consensus over how to reduce negativeimpacts and increase benefits. This process requires abandoning tradi-tional political planning approaches and adopting a more democraticplanning model. As collaboration theory indicates (Gray 1989), this re-quires the numerous stakeholders to cooperate. Jamal and Getz (1995)suggest collaboration can resolve conflicts or advance shared visions.

According to the definitions of Gray (1989) and Jamal and Getz(1995), collaboration on a more democratic planning process can bedefined as the process of joint decisionmaking among autonomous,key stakeholders of an inter-organizational, mega event domain to re-solve and manage problems and issues related to planning and host-ing. Gray (1989) suggests that the collaboration process has five keycharacteristics: stakeholders are independent; solutions emerge bydealing constructively with differences; joint ownership of decisionsis involved; stakeholders assume collective responsibility for the ongo-ing direction of the domain; and collaboration is an emergent process,where initiatives can be understood as emergent community arrange-ments through which community groups collectively cope with thegrowing complexity of hosting (Jamal and Getz 1995).

As suggested by the theory, once the bid has been won, the involve-ment and support of all stakeholders is critical, irrespective of their


previous attitudes. Indeed, community concern over both potentialbenefits and costs encourage planners to increase community involve-ment levels in the process (Haxton 1999). The need to address and re-solve societal issues raised by different community groups is likely toencourage political agents to develop collaborative strategies that im-prove the payoffs to stakeholders and reduce opposition. This ideaof collaboration also follows from the seminal work on local organiza-tions by Esman and Uphoff (1984). Their structural-reformist perspec-tive and its emphasis on a bottom-up approach to developmentsuggests a major role for the local populace in influencing decisionsthat affect them and thus helping them receive a proportion of bene-fits that might accrue. In practice, this involves sharing tasks and per-ceptions with others in the same group, or with other interest groupsat the micro and macro levels within the framework of collaborationtheory.

This discussion suggests that for a mega event to be successful, theunderstanding and participation of all stakeholders in the process iscrucial. Therefore, it is important for local governments, policymakers,and organizers to appreciate the level of community support towardthe proposed event, and to understand the basis of both support andopposition. It is also important to remember that the involvementand support of community groups and other stakeholders is likely totransform the affair into a more significant urban experience forresidents and guests alike (Hiller 1990). This is a macro variation ofstrategic management planning (Ashmos, Duchon, McDaniel andHuonker 2002) and incorporates micro application variables into amacro modeling orientation.

The Proposed Model

Figure 1 presents the model tested, which suggests that support isinfluenced by stakeholder perceptions of the potential costs and ben-efits. The model further indicates that these, in turn, are influencedby residents’ community concern, their emotional attachment to thecommunity, and their ecocentric attitude or degree of environmentalsensitivity, an especially critical issue for an intrusive Winter Olympicsvenue.

The model has its theoretical basis in social exchange theory andbuilds on previous research. Several researchers examined the per-ceived impacts on host communities (Delamere 2001; Fredline andFaulkner 2002a, 2002b; Fredline, Jago and Deery 2003). While mostof these studies focused on measuring social impacts (Delamere, Wan-kel and Hinch 2001), some of them examined variations in locals’ reac-tions towards different events (Fredline and Faulkner 2002b). Whilestudies that link perceived impacts and support are somewhat limited,many have analyzed the influence of perceived potential costs and ben-efits (Gursoy, Jurowski and Uysal 2002; Lindberg and Johnson 1997).Social exchange theory ideas are implied in their research, as it is as-sumed that individuals are likely to participate in an exchange if they

Figure 1. Proposed Model


believe they are likely to gain benefits without incurring unacceptablecosts (Homans 1974).

Theoretically, residents who view large scale tourism projects as con-tributory and believe that costs do not exceed benefits/rewards willfavor the exchange and will consequently support the process (Turner1986). However, according to this theory, the perceptions of potentialimpacts will depend, in turn, on how people evaluate the exchange inwhich they are involved. Those considering it beneficial are likely toevaluate the potential impacts differently from someone who evaluatesthe exchange as detrimental. In the context of mega event tourism, so-cial exchange theory suggests that expressed support involves a willing-ness to enter into an exchange (Gursoy and Rutherford 2004; Jurowski,Uysal, and Williams 1997).

Deccio and Baloglu (2002) proposed a social exchange theory modeldesigned to integrate factors that influence nonhost community resi-dents’ perceptions of and support for mega events. Their model envis-ages that economic gain, use of resource base, community attachment,and ecocentric attitudes all influence the perceptions of nonhost com-munity residents, and that antecedents and perceptions directly and/or indirectly influenced their support. They found that some anteced-ents not only had an indirect effect on support through their effect onthe perceptions of the opportunities and concerns, but also had a di-rect effect on support. While the model proposed by Deccio and Balo-glu (2002) provided some evidence concerning the interaction ofvarious variables and the influence of these variables on support, itwas tested on nonhost communities, which limits the applicability ofthe results.

The model proposed in this study is developed based on variablesfrom the perceived impact, on tourism support literature, and on


the Deccio and Baloglu (2002) model. It is tested on responses fromlocal residents of Salt Lake City over the 2002 Winter Olympics. Inaddition, structural modeling is used to test a new model to enableresearchers to evaluate how well the data support the model. Themodel proposed by Deccio and Baloglu (2002) used path analysis,which provides no information on how well the data support the modelsince it does not provide fit information. By contrast, the model pro-posed here will demonstrate how all factors affect the perceptions ofthe costs and benefits and will show how variables interact, also clarify-ing their direct and/or indirect causal effects on a host community’sattitudes and support. A distinct advantage of this approach is thatmore precise modeling can be used for benchmarking or barometertype applications.

Residents’ Support for Mega Events

Social exchange theory posits that residents are likely to supportmega events as long as they believe the expected benefits of develop-ment will exceed the expected costs. Mega events are likely to stimulateboth positive and negative impacts in several spheres: economic, tour-ism/commercial, physical, sociocultural, psychological, and political(Delamere 2001; Fredline et al 2003; Ritchie 1984). Unfortunately, be-cause of the intense competition to host these events, political leadersand organizers frequently ignore the negative impacts and glorify theexpected benefits. For example, the findings of a study on residents’perceptions of the impact of the 2002 World Cup Games revealed thatperceptions of impacts drastically changed after the games. Before thegames, people believed they would yield many economic and culturalbenefits for their communities, even though they were aware such ben-efits would not come without a cost. However, after the games, theyrealized that the benefits, especially the economic gains, were lowerthan they had expected (Kim, Gursoy and Lee 2006).

Even though mega sporting contests are single, short-term obtrusiveevents, they are likely to have long-term positive consequences for thecities and communities that stage them (Hiller 1990; Roche, 1994). Forinstance, they may have a lasting effect on tourism to the local commu-nity (Kang and Perdue 1994), providing opportunities for increasedinternational publicity and recognition (Jeong and Faulkner 1996),improving quality of life (Goeldner and Long 1987), generating posi-tive economic benefits, and attracting a lot of attention to the locality(Deccio and Baloglu 2002). Research suggests communities are willingto host these events mainly because of the positive economic benefitsthey can bring in the form of tax revenues, jobs, and additional sourcesof income (Getz 1997). Ritchie (1984) further adds the positive impactof enhancing awareness of the region as a domestic or internationaldestination, creating new opportunities for potential investors, andincreasing commercial activity. However, many researchers examiningresident reactions found that the most serious effects involve not onlythe economic value generated for the community, but also the changes


to the quality of residents’ life (Deccio and Baloglu 2002) and theinternational image of the host community (Jeong and Faulkner1996). Several researchers suggest that residents of communities host-ing mega events, for example, the 1988 Olympic Winter Games in Cal-gary and the 1996 Summer Olympics in Atlanta, believed that suchpositive social impacts as community pride and international recogni-tion were just as or more important than positive economic impacts(Mihalik and Cummings 1995; Ritchie and Aitken 1984, 1985).

Perhaps one of the most important perceived benefits is the facilitiescreated for the occasion but later used by locals (Mihalik and Simon-etta 1998). Residents may react positively if they believe the mega eventwill improve their recreational facilities (Allen, Hafer, Long and Per-due 1993; Kendall and Var 1984). Locals also believe these activitiesare likely to improve cultural and shopping opportunities (Jeong andFaulkner 1996), strengthen regional values and traditions, and evenlead to cultural understanding among residents and tourists (Hall1989). In addition, Deccio and Baloglu (2002) suggest proceedingsare likely to serve as catalysts for bringing attention to environmentalconcerns and thus may preserve elements of the physical landscapeand local heritage that would have otherwise been ignored. Thesefindings led to the following hypothesis:

Hypothesis 1. There is a direct positive relationship between the perceivedbenefits and the support for hosting mega events.

Although most research has focused on the positive impacts, megaevents are likely to bring both benefits and costs to the host community(Deccio and Baloglu 2002) and the former may be offset by detrimen-tal sociocultural, economic and ecological impacts, thus leading toopposition from local residents (Witt 1988). For example, they arelikely to cause price inflation and increases in local taxes to financethe facilities required to host the event. In addition, mismanagementof public funds by organizers is likely to deepen the negative economicimpacts (Deccio and Baloglu 2002). If mega events require govern-ment assistance and compete for local manpower, they may also receivenegative reactions from existing enterprises (Ritchie 1984).

Previous studies also suggest that mega events are likely to result insuch problems as traffic congestion, difficulties in law enforcement,and increased crime (Mihalik and Cummings 1995). Further, theymay damage the image of the host community or diminish its attrac-tiveness because of inadequate infrastructure, poor facilities or impro-per practices (Ritchie 1984). They can negatively influence traditionalfamily values (Kousis 1989), cause cultural commercialization (Cohen1988), and create social and cultural conflicts in the community byexacerbating differences in culture and social status and highlightingsociocultural and economic differences between hosts and tourists(Tosun 2002). Furthermore, they may be perceived to have negativeimpacts on the physical and natural environment, including pollutionand the destruction or deterioration of natural, cultural or historicalresources. Interestingly, research suggests that locals do not regard


environmental damage or cost as a major concern (Mihalik and Simon-etta 1998; Ritchie and Aitken 1984). Despite somewhat conflictingresults from earlier research, the preceding discussion has led tothe following hypothesis:

Hypothesis 2. There is a direct negative relationship between the perceived costsand the support for hosting mega events.

This study proposes that residents’ perceptions of impacts are notmutually exclusive. A change in perceptions of one type of impact islikely to influence the perceptions of other types. This suggests that ifpeople perceive benefits to be more important than costs, the percep-tions of benefits are likely to influence the perceptions of costs. In otherwords, the most salient impact is likely to influence the perception of allother impacts. It is proposed here that that there is a negative relation-ship between the costs and benefits, but it is also recognized that thedirectionality of the relationship is not known, as the relationship isnot likely to be unidirectional. Indeed, any change in how some costsor benefits are perceived may well influence how others are perceived.

Hypothesis 3. There is a direct negative relationship between the perceived costsand the perceived benefits of hosting mega events.

Attachment to the community is one determinant of support and hasbeen found to influence perception of impacts. However, previousstudies reported mixed results regarding the influence of communityattachment. For example, Um and Crompton (1987) suggest a nega-tive relationship between community attachment and the perceived im-pacts. Jurowski et al (1997) argue that ‘‘attached’’ residents are likelyto form positive perceptions of the economic and social impacts. Davis,Allen and Cosenza (1988) suggest that native residents are more posi-tive about tourism than newcomers to the community. Lankford andHoward (1994) and Gursoy et al (2002) were unable to find a clearconnection between attachment and impact perceptions, while Deccioand Baloglu (2002) found nonhost community attachment had no sig-nificant impact on perceived opportunities and support. However, theydid find community attachment had a significant influence on per-ceived concerns. McCool and Martin (1994) report a greater sense ofbelonging to a community as closely correlated to higher ratings ofboth positive and negative impacts. In the face of such conflictingresults, other hypotheses emerge.

Hypothesis 4. There is a direct relationship between community attachmentand the perceived costs.

Hypothesis 5. There is a direct relationship between community attachmentand the perceived benefits.

Another factor likely to influence perceptions of impacts is the con-cern locals feel about their community. These concerns include theenvironment, schools, crime, recreation, culture, economic develop-ment, and roads/transportation in their community. Despite somecontradictory findings, these factors have been found to influence per-


ceptions of the potential costs and benefits (Perdue, Long and Allen1990) and their support for venue development (Gursoy et al 2002).While the direction of such influences is unclear, the following hypo-theses are proposed:

Hypothesis 6. There is a direct relationship between community concern andthe perceived costs.

Hypothesis 7. There is a direct relationship between community concern andthe perceived benefits.

Ecocentrism can be defined as an individual’s orientation to soundenvironmental practices. Studies show that the level of ecocentric atti-tudes significantly affects host community reaction and their percep-tions of impacts (Jurowski et al 1997). Jurowski et al (1997) reporteda negative relationship between ecocentric attitudes and perceived im-pact factors and a nonsignificant relationship between support andecocentric values. Gursoy et al (2002) reported a negative impact onboth perceived benefits and costs relative to support and further sug-gested that this is positively influenced by the strength of the residents’ecocentrism. They argued that the positive relationship between eco-centric values and support is most likely attributable to the type ofdevelopment used to measure support. In the Jurowski study (1994),those with higher ecocentric values were shown to be more likely tofavor cultural and episode tourism than other types (attraction-basedor nature-based). Deccio and Baloglu (2002) suggested that ecocentricattitudes did not have any significant influence on impacts. Accord-ingly, the following hypotheses are considered:

Hypothesis 8. There is a direct relationship between the ecocentric attitudes oflocals and the perceived costs.

Hypothesis 9. There is a direct relationship between the ecocentric attitudes oflocals and the perceived benefits.

Research Methods

Data were collected from residents of Salt Lake City, Utah, duringthe 2002 Winter Olympics from personal interviews, using an interceptapproach. For this study, Salt Lake City was divided into four quad-rants, and interviews were conducted at strategic traffic sites in each.Participants were requested to intercept 200 residents in every quad-rant. A total of 800 Salt Lake City residents were intercepted, and420 (52.5%) agreed to participate.

For the model (Figure 1), resident support is examined as the ulti-mate dependent variable. This was measured by responses to threeitems. Residents were asked to indicate how much they would opposeor favor development of tourism services, development of informationservices, and promotion of the area as a mega event destination on a 5-point anchor scale with ‘‘strongly oppose’’ at the low end of the scaleand ‘‘strongly support’’ at the high end. Three summated scales wereused to measure the perceived benefits: positive economic impacts,


cultural, and social impacts. The perceived costs construct was alsomeasured by three summated scales: negative economic impacts, neg-ative cultural impacts, and negative social impacts. All of the items weremeasured using a 5-point anchor scale (from ‘‘strongly disagree’’ to‘‘strongly agree’’). Six items were used to measure residents’ concernfor community issues. Respondents were asked to indicate how con-cerned they were about local conditions related to environment,schools, crime, recreation, culture, economic development, androads/transportation. A 4-point anchor scale with ‘‘not at all’’ at thelow end and ‘‘very much’’ at the high end was used. Ecocentric attitudewas measured by the scale proposed by Gursoy et al (2002), based onthe new ecological paradigm scale developed by Dunlap, Van Liere,Merting, Catton and Howell (1992) using a 5-point anchor scale(‘‘strongly disagree’’ to ‘‘strongly agree’’). Three items were used tomeasure residents’ attachment to their community, as adapted fromMcCool and Martin (1994) and Goudy (1990).

Modeling Process

The analysis of the proposed model and hypothesized paths is basedon the collected data. The fits of the measurement and structural mod-els are tested using the LISREL 8 structural equation analysis package.Examination of appropriate statistics and measurement model fit indi-ces allows tests on the structural robustness.

The first statistic examined to determine the fit is the chi-square sta-tistic with associated P values. However, because of the large effect ofsample size on chi-square values (and associated P values), other indi-ces were also selected to measure the fit of the tested models based onthe recommendations of several researchers from a number of differ-ent disciplines. These selected fit indices are the goodness-of-fit index(GFI; Joreskog and Sorbom 1989), the normed-fit index (NFI; Bentlerand Bonet 1980), the parsimonious normed-fit index (PNFI; Mulaik,James, Alstine, Bennett, Lind and Stilwell 1989), the non-normed-fitindex (NNFI; Hu and Bentler 1995), the comparative fit index (CFI;Bentler 1990) and the critical N statistic (Hoelter 1983). Values ofGFI, NFI, CFI, NNFI, and PNFI range from zero to 1.00 with a valueclose to 1.00 indicating good fit (Byrne 1989; Mulaik et al 1989). Acut-off point of 200 or greater for the critical N statistic is suggestedas an indication of adequate model fit (Bollen 1989).

First, a confirmatory measurement model that specifies the positedrelations of the observed variables to the underlying constructs, withthe construct allowed to intercorrelate freely, was tested (Andersonand Gerbing 1988; Sethi and King 1994). Summated scales used inthe analysis for perceived costs and perceived benefits factors were ob-tained using SPSS. Summated scales were created for model simplicityand manageability. Before testing the overall measurement model,the unidimensionality of each construct was assessed. This procedureensures that each set of alternate indicators has only one underlyingtrait or construct in common (Sethi and King 1994). Constructs with


unacceptable fits were restructured by deleting the indicators that failedto preserve unidimensionality of the measurement (Anderson andGerbing 1988). Fit statistics, modification indices, and coefficientswere used to identify those indicators. In this analysis, assessing eachconstruct individually and deleting unacceptable indicators resultedin a decrease in the number of indicators for the constructs. Thenumber of indicators used to measure ‘‘ecocentric values’’ decreasedfrom six to three variables, and the number of indicators used tomeasure ‘‘community concern’’ decreased from seven to four variables.

The reformulated measurement model was then tested using aconfirmatory factor analysis. In an overall measurement model, theadequacy of the individual items and the composites are assessed bymeasures of reliability and validity. Three types of reliability mea-sures—composite reliability, indicator reliability, and estimated per-centage of variance extracted by each construct—are examined. Theformer, as calculated with LISREL estimates, is analogous to a coeffi-cient alpha, which shows the internal consistency of the indicatorsassessing a given factor (Hatcher 1994). Hair, Anderson, Tatham andBlack (1998) suggest that reliability scores that are between .60 and.70 are acceptable.

As shown in Table 1, the composite reliability scores of all constructsin this analysis exceeded acceptable levels. The variance-extractedestimate measures the amount of variance that is captured by a factor.The desirable level of variance captured is 50% or higher (Fornelland Larcker 1981). The table shows that the variance extracted esti-mate for each factor also exceeded the acceptable levels. Anothermeasure of reliability is the indicator reliability. However, unlike theother two reliability measures, indicator reliability does not have aspecific cut-off point in determining the acceptability of an indicator(Table 1).

Convergent validity was assessed from the measurement model bydetermining whether each indicator’s estimated pattern coefficienton its posited underlying construct factor is significant (greater thantwice its standard error). Discriminant validity was assessed for everypossible pair of constructs by constraining the estimated correlationparameter between them to 1.0 and then performing a chi-square dif-ference test on the values obtained for the constrained and uncon-strained models (Anderson and Gerbing 1988; Joreskog 1993). Asignificantly lower chi-square value in an unconstrained model indi-cates that discriminant validity is achieved.

Results showed all constructs to have both convergent and discrimi-nant validity. The overall fit of this final measurement model wasv2ð133Þ ¼ 210:84 (p = 0.0); GFI = .95; AGFI = .93; NFI = .91; NNFI = .94;

CFI = .96, and PGFI = 0.66; PNFI = .69 and critical N = 343.11. The indi-cators of residuals, RMR (root mean square), standardized RMR andRMSEA (root mean square error of approximation) were .029, .039and .037, respectively.

After making sure the measurement model was acceptable, theproposed structural equation model was tested. Figure 1 presents the

Table 1. Measurement Scale Properties (N = 420)

Constructs and Indicators CompletelyStandardized

Loadings

IndicatorReliability

ErrorVariance

Community Attachment 0.60b 0.63a

How much do you feel at home in thiscommunity?

0.57 0.32 0.68

What interest do you have in knowing whatgoes in this community?

0.52 0.27 0.73

Assume that you have been livingin this communityfor a while now. Suppose that forsome reason youhad to move awayfrom this community, how sorryor pleased would you be to leave

0.7 0.49 0.51

Ecocentric Attitude 0.67b 0.72a

When human interfere with natureit often producesdisastrous consequences

0.67 0.45 0.55

The balance of nature is strongenough to cope withthe impacts of modern industrialnations. (RECODED)

0.48 0.23 0.77

If things continue on their presentcourse, we will soonexperience a major ecologicalcatastrophe

0.86 0.74 0.26

Community Concern 0.70b 0.79a

Crime 0.65 0.42 0.58Recreation 0.81 0.66 0.34Culture 0.74 0.55 0.45Roads/Transportation 0.58 0.34 0.66Perceived Cost of Mega Events 0.59b 0.62a

Negative economic impacts 0.52 0.27 0.73Negative cultural impacts 0.78 0.61 0.39Negative social impacts 0.46 0.21 0.79Perceived Benefits of Mega Events 0.60b 0.63a

Positive economic impacts 0.48 0.23 0.77Positive cultural impacts 0.72 0.52 0.48Positive social impacts 0.59 0.35 0.65Support for Mega Events 0.63b 0.66a

Visitor services (for example,hotels, restaurants)

0.56 0.31 0.69

Information services for visitors(such as maps and guidebooks)

0.61 0.37 0.63

Promotion of the area as a mega eventdestination (such as television advertisingor brochures)

0.71 0.50 0.50

a Composite reliability of each construct. b Variance extracted estimate.



proposed model. The fit indices of the proposed model suggest that itwas acceptable. The overall fit of the structural model wasv2ð135Þ ¼ 228:51 (p = 0.0); GFI = .95; AGFI = .92; NFI = .90; NNFI = .93;

CFI = .95, and PGFI = 0.67; PNFI = .70 and critical N = 320.88. The indi-cators of residuals, RMR, standardized RMR and RMSEA were .031,.043 and .041, respectively.

Attitudes towards Mega Events

An analysis of the estimated standardized path coefficients revealedthe significance, strength, and direction of each hypothesized relation-ship. Seven of the nine hypothesized paths were statistically significantat the .05 probability level. Two of the proposed hypotheses wererejected. First, the perceived cost was not significant in its effect on sup-port (b = �.06; t = �.84). Consequently, Hypothesis 3, a negative rela-tionship between the perceived costs and support, was notsupported. This finding contradicts the conclusion of many early stud-ies, which suggested the perceived cost negatively relates to support(Keogh 1990). But these studies did not consider the more conspicu-ous or obtrusive type of tourism. Kim et al (2006), for example, suggestthat communities that host mega events often ignore negative impactsprior to hosting, while glorifying the expected benefits. Deccio andBaloglu (2002) also reported an insignificant relationship betweenperceived costs and support. Consequently, the results of this studysupport the findings and arguments of others that the insignificantimpact of perceived cost may be explained by the type or level ofdevelopment in the community in which the study took place (Allenet al 1993; Gursoy et al 2002).

Second, Hypothesis 4, proposing a direct relationship between com-munity attachment and perceived costs, was not supported (b = �.05;t = 0.72). This finding contradicts the findings of other studies (Deccioand Baloglu 2002; Gursoy et al 2002) that suggest the level of commu-nity attachment is likely to have a significant effect on the evaluation ofthe costs. The type of development, an obstructive, highly visible event,may explain the lack of a relationship. A mega event, an obtrusive formof tourism, is a single event that usually generates long-term profoundimpacts (Mihalik and Simonette 1998). Since it is a single, short-termevent, residents may not be so concerned with its costs and may thusoverlook those associated with the event.

As hypothesized, evidence was found to support the direct positiverelationship between perceived benefits and support (b = .43;t = 5.35). This finding is consistent with previous studies and social ex-change theory that suggest that perceived benefits positively affect thelevel of host community support for hosting a mega event (Deccio andBaloglu 2002; Getz 1997). Support was also found for the hypothesispredicting a negative relationship between perceived costs and per-ceived benefits (b = �.27; t = 3.02). However, this finding suggests onlythat there is a negative relationship between those two; it does notclarify or make any claim about the directionality of the relationship.


The insignificant relationship between costs and support may also beexplained by this finding. As noted earlier, perceptions of impactsare not mutually exclusive; a change in perceptions of one type is likelyto influence the perceptions of other types of impacts. This suggeststhat if residents place more importance on benefits, they may overlookthe costs associated with hosting the event.

The hypothesis predicting a direct relationship between level of com-munity attachment and perceived benefits was supported (b = .11;t = 2.57). This finding is also consistent with previous studies (Deccioand Baloglu 2002), which suggest that residents who are highly at-tached to their community are more likely to view the mega event ascreating benefits for the local community. This finding may also ex-plain the insignificant relationship between community attachmentand costs. As discussed earlier, residents may be extremely willing toparticipate in an exchange because of anticipated benefits, and maytherefore overlook the costs.

The hypothesized effects of ecocentric attitudes on the perceivedcosts and benefits were found to be significant (b = .26 and .17;t = 2.36 and 3.16 respectively), supporting previous work (Gursoyet al 2002). This indicates that high ecocentric values are associatedwith high perceptions of costs and benefits. However, for residents withhigh ecocentric values, the perceptions of benefits are likely to be lesssalient than the perceptions of costs.

Finally, the hypothesis predicting a direct relationship between com-munity concern and the perceived costs, and the hypothesis predictinga direct relationship between community concern and the perceivedbenefits, were supported (b = �.22 and .38; t = 2.51 and 4.62 respec-tively). This suggests that people who are highly concerned with com-munity issues are likely to see mega events as generating long-termprofound impacts on their communities, both positive and negative,as indicated by several researchers (Mihalik and Cummings 1995;Mihalik and Simonette 1998). Overall, the structural model explained31% of the variance in the ultimate dependent variable, support formega events.

CONCLUSION

Structural modeling was used to test a new model to enable theresearchers to evaluate how well the data support the model. It is un-like work by Deccio and Baloglu (2002), which used path analysis,which does not provide information on how well the data supportthe model with ‘‘fit’’ information. The model proposed in this studyhas demonstrated how every factor affects the perceptions of the costsand benefits and identifies the interactions among the variables. Themodel utilized clarified direct and/or indirect causal effects on a hostcommunity’s reaction to and support for mega events. The advantageis that this more precise modeling can be used for benchmarking orbarometer-type applications.


Although it is more difficult to implement, communities are morefrequently adopting a more collaborative decision making approachto mega event planning which combines both technical rationalityand participatory democracy in the overall planning process (Haxton1999; Jafari 1990). Communities are slowly abandoning the hallmarkdecisionmaking (political planning) approach. The successful imple-mentation of the democratic approaches, as in Toronto’s Bid for1996 Summer Olympic Games and the Calgary Olympics, is fueling thisnew trend. Politicians and organizers are starting to understand thevalue of locals’ involvement and support for hosting mega events.Recently, residents’ involvement and support has become even moreimportant for communities planning to bid for Olympic venues. TheInternational Olympic Committee has stated that it will no longer en-dorse or allow a host city to fund future games solely from private re-sources. The committee’s need for more public funding is likely torequire an increased of level local involvement and support to endorsean increase in taxation to fund the massive and expensive infrastruc-ture projects that will be needed to host Olympic projects (Mihalik2000). This should encourage political agents to collaborate more withother stakeholders and solicit inputs from community groups beforethe bidding to stimulate public debate and more community involve-ment (French and Disher 1997).

The emerging collaborative decisionmaking trend makes hostinglarge venues a more complex phenomenon due to the number ofstakeholders involved and affected by the process. While successfulhosting depends on how well it is organized, the quality of the infra-structure and facilities, and attractions and services, it also requires in-creased support and hospitality from locals. Once a community decidesto be a host, the quality of life of the locals is likely to be affected.Anger, apathy, or mistrust by the local population will ultimately beconveyed to the tourists and is likely to result in an unsatisfactory expe-rience (Gursoy et al 2004).

Support and involvement of the locals are necessary for three impor-tant reasons. First, they are often asked to vote for tax increases to sup-port infrastructure and facilities. Second, a friendly and hospitablelocal population is critical in transforming a mega event into an urbanfestival to provide a significant experience for residents and guestsalike (Hiller 1990). Three, local support and involvement are likelyto increase the longevity of positive impacts on the local community.Knowledge of the factors affecting host community support and theinterplay of those factors may enable communities to assess the levelof support by the stakeholders as the collaboration proceeds. This willavoid committing large amounts of financial and other resourcesbefore community concerns are considered.

Communities may utilize the modeling in this study as a prototype.In addition, findings could help develop communication strategiesthat deal with specific issues raised by various community groups andstakeholders. In order to assess the level of support, communities needa better understanding of what is important to the stakeholders. Once


they identify what is significant, they may apply the principles of inter-nal marketing to solicit support from the stakeholders.

The results suggest communities should evaluate the level of con-cern stakeholders have about the community, their attitudes towardthe environment, and their attachment to the community prior to pro-posing a plan to host a mega event. The greater the concern for thecommunity, the more support a mega event-hosting proposal is likelyto attain. Internal marketing techniques designed to inform locals ofthe benefits they are likely to receive may help win their support. Thisstudy indicates that residents who expressed a high level of attachmentto their communities are more likely to view hosting a mega event asbeneficial. This suggests that such residents can then be recruited assupporters of the event and thus increase local involvement and collab-oration in the process.

As suggested by previous studies, a strong ecocentric value within thecommunity will not necessarily result in opposition. However, it is sug-gested here that residents with high ecocentric values are likely to paymore attention to costs. Therefore, communities need to recognizethat it will be more difficult to promote the benefits of hosting to stake-holders if a large section of them affirm strong ecocentric values. Sen-sitivity to environmental concerns is crucial to win the support of localswith ecocentric attitudes. Collaboration with such groups and theirinvolvement in the process may increase support, and proposals todevelop conservation and preservation programs may help to easethe concerns of the more ecocentric residents.

It is readily apparent from the results that in the case of obtrusivetourism venues, the perceived impact of costs does not have thesame implications as it might from lesser venues. This mitigationmay be explained by the fact that these events are world-class, uniqueevents. As a consequence, residents may perceive that the benefits re-ceived outweigh the costs of being a host. Findings also suggest thatraising funds for obtrusive venues may be more effective than for lessobtrusive developments. Since these events are broadcast worldwide,locals may be more willing to spend money on them to make the com-munity appear impressive. This certainly allows host locations moreleeway in budgetary matters. Residents with increased pride and self-esteem, perhaps associated with the attention the community receives,tend to accommodate the higher costs of such developments. Thiswould appear to be the case in this study, as the traditional measureof community concern effect was negated by other interveningvariables.

Previous studies suggest that a temporal effect may also be present.Communications, perceptions, and visual impacts may be quite differ-ent before, during, and after the mega event. For example, Mihalik(2000) reports that residents’ perceptions of the impacts of the 1996Atlanta Olympics Games changed over time. While support remainedstrong, people became increasingly concerned about the negative im-pacts (perceived liabilities). Similar findings were reported by Kimet al (2006) that local residents’ perceptions of the impacts changed


drastically after the event was held. Before it started, they had highexpectations about the economic and cultural benefits their communi-ties would receive, though they were aware that such advantages wouldnot be cost free. However, they later realized that the benefits gener-ated, especially economic, were lower than they had anticipated. Thisstudy did not examine the temporal effects, but the effect of time onthe proposed construct and the hypothesized relationships is certainlya subject of future research. In addition, other constructs were not con-sidered in this modeling process. In particular, participation in recre-ation and leisure pursuits of the residents could also have a directpositive effect on perceived costs and support for tourism develop-ment. A Winter Olympic venue may not provide as much recreationand leisure pursuit support as the more extensive Summer Olympicsor Soccer World Cup venues.

One final aspect not examined was the level of community involve-ment in the preliminary bidding and planning process of hosting theSalt Lake City Winter Olympics. In other words, what respondents actu-ally knew about the Olympics, how they were informed, the debatesthat occurred in the community and the media, and their pre-formedopinions on hosting the Winter Olympic Games were not assessed. Theinclusion of these factors as moderating variables might possibly havechanged the estimated coefficients. Future studies should includequestions about the level of community involvement and local know-ledge about the mega event to assess the moderating affect of thesefactors on estimated coefficients and on support.

REFERENCES

Allen, L., H. Hafer, R. Long, and R. Perdue1993 Rural Residents’ Attitudes toward Recreation and Tourism Development.

Journal of Travel Research 31(4):27–33.Anderson, J., and D. Gerbing

1988 Structural Equation Modeling in Practice: A Review and RecommendedTwo-Step Approach. Psychological Bulletin 103:411–423.

Ashmos, D., D. Duchon, R. McDaniel, and J. Huonker2002 What A Mess! Participation as a Simple Managerial Rule to ‘‘Complexify’’

Organizations. Journal of Management Studies 39:189–206.Bentler, P.

1990 Comparative Indexes in Structural Models. Psychological Bulletin107:238–246.

Bentler, P., and D. Bonett1980 Significance Test and Goodness of Fit in the Analysis of Covariance

Structures. Psychological Bulletin 88:591–606.Bollen, K.

1989 Structural Equations with Latent Variables. Wiley: New York.Byrne, N.

1989 A Primer of LISREL: Basic Applications and Programming for Confir-matory Factor Analytic Models. New York: Springer-Verlag.

Cohen, E.1988 Tourism and Aids in Thailand. Annals of Tourism Research 15:467–486.

Davis, D., J. Allen, and R. Cosenza1988 Segmenting Local Residents by their Attitudes, Interests, and Opinions

toward Tourism. Journal of Travel Research 27(3):2–8.


Deccio, C., and S. Baloglu2002 Nonhost Community Resident Reactions to the 2002 Winter Olympics:

The Spillover Impacts. Journal of Travel Research 41:46–56.Delamere, T.

2001 Development of a Scale to Measure Resident Attitudes toward the SocialImpacts of Community Festivals, part II: Verification of the scale. EventManagement 7:25–38.

Delamere, T., L. Wankel, and T. Hinch2001 Development of a Scale to Measure Resident Attitudes toward the Social

Impacts of Community Festivals. Part I: Item Generation and Purification ofthe Measure. Event Management 7:11–24.

Dunlap, R., K. Van Liere, A. Mertig, W. Catton, and R. Howell1992 Measuring Endorsement of an Ecological Worldview: A Revised NEP

scale. Presented at the Annual Meeting of the Rural Sociological Society.Esman, M., and N. Uphoff

1984 Local Organizations: Intermediaries in Rural Development. Ithaca:Cornell University Press.

Fornell, C., and D. Larcker1981 Evaluating Structural Equation Models with Unobservable Variables and

Measurement Error. Journal of Marketing Research 18(1):39–50.Fredline, E., and B. Faulkner

2002a Residents’ Reactions to the Staging of Major Motorsport Events withintheir Communities: A Cluster Analysis. Event Management 7:103–114.

2002b Variations in Residents’ Reactions to Major Motorsport Events: WhyResidents Perceive the Impacts of Events Differently. Event Management7:115–125.

Fredline, L., L. Jago, and M. Deery2003 The Development of a Generic Scale to Measure the Social Impacts of

Events. Event Management 8:23–37.French, S., and M. Disher

1997 Atlanta and the Olympics: A One-Year Retrospective. Journal of theAmerican Planning Association 63:379–392.

Getz, D.1991 Festivals, Special Events, and Tourism. New York: Van Mostrand Reinhold.1997 Event Management and Event Tourism. New York: Cognizant

Communication.Goeldner, R., and P. Long

1987 The Role and Impact of Mega-Events and Attractions on TourismDevelopment in North America. Proceedings of the 37th Congress of AIEST28, 119–131.

Goudy, W.1990 Community Attachment in Rural Region. Rural Sociology 55:178–198.

Gray, B.1989 Collaborating: Finding Common Ground for Multiparty Problems. San

Francisco: Jossey-Bass.Gursoy, D., and D. Rutherford

2004 Host Attitudes toward Tourism: An Improved Structural Model. Annalsof Tourism Research 31:495–516.

Gursoy, D., C. Jurowski, and M. Uysal2002 Resident Attitudes: A Structural Modeling Approach. Annals of Tourism

Research 29:79–105.Hair, J., R. Anderson, R. Tatham, and W. Black

1998 Multivariate Data Analysis (fifth ed.). Upper Saddle River: Prentice Hall.Hall, C.

1989 Hallmark Tourist Events: Analysis, Definition, Methodology and Review.In The Planning and Evaluation of Hallmark Events, G. Syme, B. Shaw, D.Fenton and W. Mueller, eds., pp. 3–40. Sydney, Australia: Avebury.

Hatcher, L.1994 A Step-by-Step Approach to Using the SAS System for Factor Analysis and

Structural Equation Modeling. Cary NC: SAS.


Haxton, P.1993 A Post Event Evaluation of the Social Impacts and Community Percep-

tions of Mega Events. Honors Thesis. Townsville QLD: James Cook University.1999 Community Involvement and the Olympic Games: A Review of Related

Research. 7th International Post Graduate Seminar on Olympic Studies,Greece.

Hernandez, S., J. Cohen, and H. Garcia1996 Residents’ Attitudes towards an Instant Resort Enclave. Annals of

Tourism Research 23:755–779.Hiller, H.

1990 The Urban Transformation of a Landmark Event: The 1988 CalgaryWinter Olmpics. Urban Affairs Quarterly 26:118–137.

Hoelter, J.1983 The Analysis of Covariance Structures: Goodness-of-Fit Indices. Socio-

logical Methods and Research 11:325–344.Homans, G.

1974 Social Behavior: Its Elementary Forms (Revised ed.). New York: HarcourtBrace Jovanovich.

Hu, L., and P. Bentler1995 Evaluating Model Fit. In Structural Equation Modeling: Concepts, Issues

and Applications, R. Hoyle, ed., pp. 76–99. Thousand Oak: Sage.Jafari, J.

1990 Research and Scholarship: The Basis of Tourism Education. Journal ofTourism Studies 1(1):33–41.

Jamal, T., and D. Getz1995 Collaboration Theory and Community Tourism Planning. Annals of

Tourism Research 22:186–204.Jeong, G., and B. Faulkner

1996 Resident Perceptions of Mega-Event Impacts: The Taejon InternationalExposition Case. Festival Management & Event Tourism 4(1):3–11.

Joreskog, K., and D. Sorbom1989 LISREL 7: User’s Reference Guide (first ed.). Chicago: Scientific

Software.1993 LISREL VIII Manual. Mooresville, IN: Scientific Software.

Jurowski, C.1994 The Interplay of Elements Affecting Host Community Resident Attitudes

toward Tourism: A Path Analytic Approach. PhD dissertation. VirginiaPolytechnic Institute and State University.

Jurowski, C., M. Uysal, and R. Williams1997 A Theoretical Analysis of Host Community Resident Reactions to

Tourism. Journal of Travel Research 36(2):3–11.Kang, Y., and R. Perdue

1994 Long-Term Impact of a Mega-Event on International Tourism to theHost Country: A Conceptual Model and the Case of the 1988 Seoul Olympics.Journal of International Consumer Marketing 6:205–226.

Kendall, K., and T. Var1984 The Perceived Impact of Tourism: The State of the Art. Vancouver:

Simon Fraser University.Keogh, B.

1990 Resident and Recreationists’ Perceptions and Attitudes with Respect toTourism Development. Journal of Applied Recreation Research 15(2):71–83.

Kim, J., D. Gursoy, and S. Lee2006 The Impact of the 2002 World Cup on South Korea: Comparisons of Pre-

and Post-Games. Tourism Management 27:86–96.Kousis, M.

1989 Tourism and the Family in a Rural Cretan Community. Annals ofTourism Research 16:318–332.

Lankford, S., and D. Howard1994 Developing a Tourism Attitude Impact Scale. Annals of Tourism

Research 21:121–139.


Lenskyj, H.1992 More than Games: Community Involvement in Toronto’s Bid for the

1996 Summer Olympics. In Proceedings of the First International Symposiumfor Olympic Research, R. Barney and K. Meier, eds., 78–87. London: OntarioCenter for Olympics Studies, University of Western Ontario.

Lindberg, K., and R. Johnson1997 Modeling Resident Attitudes toward Tourism. Annals of Tourism

Research 24:402–424.McCool, S., and S. Martin

1994 Community Attachment and Attitudes toward Tourism Development.Journal of Travel Research 32(3):29–34.

Mihalik, B.2000 Host Population Perception of the 1996 Atlanta Olympics: Support,

Benefits, and Liabilities. Tourism Analysis 5:49–53.Mihalik, B., and P. Cummings

1995 Host Perceptions of the 1996 Atlanta Olympics: Support, Attendance,Benefits and Liabilities. In Travel and Tourism Research Association 26thAnnual Proceedings, pp. 397–400.

Mihalik, B., and L. Simonette1998 Resident Perceptions of the 1996 Summer Olympic Games: Year II.

Festival Management and Event Tourism 5(1):9–19.Mulaik, S., L. James, J. Alstine, N. Bennett, S. Lind, and C. Stilwell

1989 Evaluation of Goodness-of-Fit indices for Structural Equation Models.Psychological Bulletin 10:430.

Perdue, R., P. Long, and L. Allen1990 Resident Support for Tourism Development. Annals of Tourism

Research 17:586–599.Ritchie, J.

1984 Assessing the Impact of Hallmark Events: Conceptual and ResearchIssues. Journal of Travel Research 22(1):2–11.

Ritchie, J., and C. Aitken1984 Olympulse I: The Research Program and Initial Results. Journal of Travel

Research 22(1):17–25.1985 Olympulse II: Evolving Resident Attitudes toward the 1998 Olympic

Winter Games. Journal of Travel Research 23(1):28–33.Roche, M.

1994 Mega-events and Urban Policy. Annals of Tourism Research 21:1–19.

Sethi, V., and W. King1994 Development of Measures to Assess the Extent to which an Information

Technology Application Provides Competitive Advantage. Management Sci-ence 40:1601–1624.

Tosun, C.2002 Host Perceptions of Impacts: A Comparative Tourism Study. Annals of

Tourism Research 29:231–245.Turner, J.

1986 The Structure of Sociological Theory. Chicago: The Dorsey Press.Um, S., and J. Crompton

1987 Measuring Resident’s Attachment Levels in a Host Community. Journalof Travel Research 26(2):27–29.

Veal, A.1994 Leisure Policy and Planning. Harlow: Lomhman.

Webb, E., D. Campbell, R. Schwartz, and L. Sechrest1971 Unobtrusive Measures: Nonreactive Research in the Social Sciences.

Chicagol: Rand McNally.Williams, J., and R. Lawson

2001 Community Issues and Resident Opinions of Tourism. Annals ofTourism Research 28:269–290.

Witt, S.1998 Mega-Events and Mega-Attractions. Tourism Management 9:76–77.


Submitted 5 April 2004. Resubmitted 29 November 2004. Resubmitted 7 May 2005.Resubmitted July 28 2005. Resubmitted 14 September 2005. Final version 4 October 2005.
Accepted 11 December 2005. Refereed anonymously. Coordinating Editor: David H.Harrison

hosting mega events: modeling locals’ support

Documents