iod

cienyi U

ip b< Centralization is positively related to OCB, and negatively related to DWB.

In todays struggling global economy, it is not enough foremployees to merely do their jobs. Many businesses expect orga-nizational members to not only complete their required duties, butalso proactively assist their colleagues. Therefore, many

a centralized organization are negatively related to OCBs (DeGroot& Brownlee, 2006; Raub, 2008). Moreover, centralizationwill resultin job dissatisfaction and employee DWB may be evoked (Judge,Scott, & Ilies, 2006; Willem, Buelens, & Jonghe, 2007). In addition,a more centralized organization may elicit work disinterest fromemployees and reduce their work enthusiasm. There may befunctional aspects of employee disinterest, however, and they mayturn to their coworkers for conversation (Matheson, 2007) andthereby strengthen workplace friendships (Sias & Cahill, 1998). If

* Corresponding author. Tel.: 886 4 22196512; fax: 886 4 22196511.E-mail addresses: chyan@nutc.edu.tw (C.-H. Yen), serena.alex@msa.hinet.net

(H.-Y. Teng).1

Contents lists available at

Tourism Ma

ls

Tourism Management 36 (2013) 401e410Tel.: 886 4 22023427; fax: 886 5 2732923.One of the challenges of hospitality industry management is toachieve a balance between the need for efciency versus that forcustomization. Efciency requires rules, standardization, and oftenformalized, mechanistic ways of doing work. In contrast, custom-ization requires empowerment, openness, and more organicorganizational designs (gaard, Marnburg, & Larsen, 2008). Orga-nizational theories generated from Western countries remaina challenge in Oriental countries such as Taiwan, due to thedifferences in cultures (Shenkar & Von Glinow, 1994). Therefore, itis important to examine the problems of using the same manage-ment practices across cultures in the hospitality industry.

behaviors such as organizational citizenship behavior (OCB).However, research has shown that deviant workplace behavior(DWB) is fairly prevalent (Harper, 1990; McGurn, 1988). Whenemployees participate in DWB, the behavior can have devastatingeffects on the organization or its members (Robinson & Bennett,1997). As a result, researchers have been encouraged to developa deeper understanding of the variables associated with OCBs(Huang, Iun, Liu, & Gong, 2010; Kim, ONeill, & Cho, 2010) and DWBs(Avey, Luthans, & Youssef, 2010; Bowling, 2010).

Prior research has indicated that hotel companies have thecharacteristics of high levels of centralization (gaard et al., 2008).Several researchers found that employees perceptions of1. Introduction organizational scholars are focusing on employees positivea r t i c l e i n f o

Article history:Received 16 March 2012Accepted 3 October 2012

Keywords:CentralizationOrganizational citizenship behaviorDeviant workplace behavior0261-5177/$ e see front matter 2012 Elsevier Ltd.http://dx.doi.org/10.1016/j.tourman.2012.10.003a b s t r a c t

Worker deviant behavior, and its opposite, organizational citizenship are two key aspects of employeediscretionary behavior, and are important in hotel operations. The purpose of this study is to examine theinuence of centralization on organizational citizenship behaviors (OCB) and deviant workplacebehaviors (DWB). Survey data from 318 employees in Taiwans hotel industry indicate that centralizationis positively related to OCB, and negatively related to DWB. Moreover, procedural justice partiallymediated the relationship between centralization and OCB/DWB. These results have importance formanagement and the paper concludes by discussing the implications of the results.

2012 Elsevier Ltd. All rights reserved.< Procedural justice partially mediates the relationship between centralization and OCB/DWB.The effect of centralization on organizatworkplace behavior in the hospitality in

Chang-Hua Yen a,*, Hsiu-Yu Teng b,1

aDepartment of Leisure and Recreation Management, National Taichung University of SbGraduate Institute of Recreation, Tourism and Hospitality Management, National Chia

h i g h l i g h t s

Mathey have stronger friendships with coworkers, theymight bemoreinclined to help those coworkers (OCB) and less inclined to harmthem (DWB).

The above ndings indicate that centralization has signicantinuence on OCB and DWB. However, whether the impact ispositive or negative is obviously inconclusive from the abovedeductions. Therefore, the purpose of this research is to examinethe effect of centralization on OCB and DWB. Moreover, based onthe social exchange theory (Blau, 1964), previous literature indi-cated that procedural justice is an important antecedent of OCB(Colquitt, Conlon, Wesson, Porter, & Ng, 2001) and DWB (Dineen,Lewicki, & Tomlinson, 2006; Skarlicki & Folger, 1997). This studyalso examines the mediating effect of procedural justice on therelationship between centralization and OCB and DWB.

2. Literature review and research hypotheses

2.1. Organizational structure in the hospitality industry

Several researchers considered that hotel managers are inu-enced by a traditional idea of leadership and management(Pittaway, Carmouche, & Chell, 1998; Tracey & Hinkin, 1996), andthat employees follow particular systems and instructions given bytheir superiors (gaard et al., 2008). Hotel companies are charac-terized by higher centralization. The relative degree of centraliza-tion is signied by the hierarchy of authority and by employeeslack of participation in decision-making. Past researchers havedescribed the practices and policies in hotel companies as archaicand inexible (Tracey & Nathan, 2002), and accepting the way wedo things around here may be the value that is deeply ingrained inthe culture of the hospitality industry (Raub, 2008). Thus, ingeneral, hospitality organizations represent relatively centralizedorganizations, needing routines and formal systems to coordinatetasks and secure efcient accomplishment of organizationalobjectives (gaard et al., 2008).

However, Pavia and Pilepic (2010) indicated that traditionalorganizational units in hotels are changed into process teams, andtasks become more exible. The purpose of employees is no longerto carry out the assignments given by supervisors, but rather tofulll customer needs and enhance customer satisfaction.Empowerment has been described as an organizational changestrategy for the hospitality industry (Erstad, 1997). Decision-making becomes a part of each process and takes place at the siteof each process in hotels, meaning that all employees have theautonomy and responsibility for making decisions (Pavia & Pilepic,2010). Accordingly, both centralized and decentralized structuresare prevalent in the hospitality industry.

2.2. Organizational citizenship behavior

OCB is dened as individual behaviors that are discretionary andnot rewarded directly by the organization (Organ, Podsakoff, &MacKenzie, 2006). Williams and Anderson (1991) created a two-dimensional conceptualization of OCB consisting of OCB directedto individuals (OCBI) and OCB directed to the organization (OCBO).OCBI immediately benets particular individuals within the orga-nization. Such behavior may include helping colleagues who haveheavier workloads (Ertrk, 2007). OCBO benets the organizationas a whole, such as punctuality, having a positive attitude, andmaking suggestions for the organizations improvement (Gilbert,Laschinger, & Leiter, 2010). OCB is important for organizationsbecause it facilitates the accomplishment of organizational goalsand enhances organizational performance (Fisher, McPhail, &Menghetti, 2010). Moreover, OCB can enhance customer satisfac-

C.-H. Yen, H.-Y. Teng / Tourism402tion (Podsakoff, Whiting, Podsakoff, & Blume, 2009). Therefore,OCB has become critical in todays corporate world, where orga-nizations have to be increasingly effective to survive.

2.3. Deviant workplace behavior

Robinson and Bennett (1995) dened employee deviance asvoluntary behavior that violates important organizational normsand threatens the well-being of organizations, its members, orboth. Bennett and Robinson (2000) identied two facets of DWBconsisting of DWB directed at individuals (DWBI) and DWBdirected at the organization (DWBO). DWBI takes the form ofbehaviors directed at specic individuals of the organization andcan include abuse, rudeness and physical assault (Mulki, Jaramillo,& Locander, 2006; Robinson & Bennett, 1995). DWBO is directedagainst the organization and includes such actions as stealing andwithholding effort (Colbert, Mount, Harter, Witt, & Barrick, 2004).DWB produces organizational losses estimated to reach up to $200billion annually in the United States (Harris & Ogbonna, 2006). Thisjusties the importance of studying DWBs in an organization.

2.4. Cross-cultural differences

Based on Hofstedes (1980) research, the dominant culturalpattern in Western countries is individualism, whereas that inAsian countries such as Taiwan is collectivism. Taiwan holdsdistinct cultural values that reect the Confucian ethics. Confu-cianism places high value on total loyalty to a hierarchical structureof authority: a code of dened conduct between subordinates andmanagers, and trust among friends (Lee & Liu, 2007). These rela-tionships are based on mutual and complementary obligations(Hofstede, 1991). Individuals with an adherence to Confucianismbelieve that social harmony and common interests are moreimportant than individual interests and enjoyment. Also, peopleare more socio-centric and have an interdependent view of the self(Markus & Kitayama, 1991). In addition, guanxi as connections orrelationships plays a critical role in the Taiwanese society. Guanxirefers to developing interpersonal relationship through formal orinformal exchange and progressing mutual trust network withconsensus of benet and restraint (Lin & Ho, 2010). In business,guanxi generally involves a hierarchical network of interpersonalrelationships embedded with mutual obligations to exchangefavors or affection (Wong & Tam, 2000).

Previous literature found that culture differences have impactson the OCB (Cohen & Avrahami, 2006) and DWB (Jackson, Colquitt,Wesson, & Zapata-Phelan, 2006). Taormina and Gao (2010) identifythree components of guanxi behavior: seeking assistance fromfamily, maintaining friendships by providing help to friends, anddoing favors for associates. Lin and Ho (2010) found that peoplewith collectivism inclination engage in the OCB more positivelysince group harmony and unity are mainly focused in collectivism.On the contrary, people with individualism are more likely toconsider themselves as independent members of a group. Jacksonet al. (2006) also stated that collectivists performed their grouptasks better, contributed more discretionary citizenship, and wereless likely to engage in deviant behaviors. Therefore, nationalculture might encourage or inhibit OCB and DWB.

2.5. Centralization and organizational citizenship behavior

Centralization refers to the concentration of power or decision-making authority in an organization (Schminke, Cropanzano, &Rupp, 2002). High centralization inhibits interactions amongorganizational members (Gold, Malhotra, & Segars, 2001). Highlycentralized organizations prevent communication (Pertusa-Ortega,

nagement 36 (2013) 401e410Zaragoza-Sez, & Claver-Corts, 2010) and reduce intrinsic

ZMadaHighlight

negatively related to OCB. Conversely, past research indicated thatparticipation in decision-making can lead to engagement in OCBs,

tinguishing between DWBO and DWBI (Bennett & Robinson, 2000;Robinson & Bennett, 1997). If the organization is the cause of themistreatment, then deviance will be most likely directed against

Mansuch as helping new members of the work group (Porter, Lawler, &Hackman, 1996). Centralization is negatively associated withempowerment. Empowerment enhances feelings of self-efcacyamong organizational members, and organizational membersmay reciprocate by performing OCBs (Wat & Shaffer, 2005). VanYperen, van den Berg, and Willering (1999) also found that themore employees feel that they participate in decision-making, themore they feel supported by their supervisor, which is accompa-nied by the display of more OCBs. Moreover, Gilbert et al. (2010)suggested that empowerment is more strongly related toemployee discretionary behaviors directed at the organization thantoward the individual. OCBO is more likely to be a direct function ofwhat employees think about their work characteristics. OCBI mightindeed reect a natural expression of employees affect at workrather than their deliberate attempt to restore the balance with theorganization. According to the above research, it could be theorizedthat centralization may be negatively related to OCBI and OCBO.Accordingly, the following hypotheses are presented:

H1a. Centralization is negatively related to OCBI.

H1b. Centralization is negatively related to OCBO.

Alternatively, centralization may indirectly enhance OCB.Matheson (2007) stated that centralized organizations may elicitemployees work disinterest and lower work enthusiasm. Central-ization increases employees alienation by limiting their autonomyand their selection of goals. Hence, centralization contributes tofeelings of powerlessness (Conger & Kanungo,1988) and employeeswould transfer their focus to other activities, such as conversation,which will in turn enhance workplace friendship (Sias & Cahill,1998). Strong friendship ties lead to reciprocity and employeessocial exchange toward their friends. Whenworkplace friendship isincreased, employees will be more willing to assist colleagues(Bowler & Brass, 2006). In addition, highly cohesive groups arelikely to have a strong sense of social identity and belonging thatcan increase organizational members desire to help one another(Van Dyne, Graham, & Dienesch, 1994) and to engage in OCBs.Hence, group cohesiveness is positively related to OCBs (Tan & Tan,2008). In sum, we can theorize that higher centralization of orga-nizational structure may enhance employee OCBI and OCBO.

According to the above literature review and since the effect ofcentralization on OCB is somewhat ambivalent, this researchproposes the following contrasting hypotheses:

H2a. Centralization is positively related to OCBI.

H2b. Centralization is positively related to OCBO.

2.6. Centralization and deviant workplace behavior

Highly centralized organization discourages employees fromexertingmore efforts in achieving organizational goals (Organ et al.,2006). Past literature found that centralizationwill result in greaterdissatisfaction among employees (Willem et al., 2007). The socialmotivation and employee satisfaction (Zheng, Yang, & McLean,2010), which is negatively related to OCBs (Lapierre & Hackett,2007; Nadiri & Tanova, 2010). Raub (2008) also indicated thatcentralized structure which restrict employees margin of controlwill have a negative impact on their propensity to display OCB.Therefore, employees perceptions of a centralized organization are

C.-H. Yen, H.-Y. Teng / Tourismexchange theory predicts that individuals who perceive that theythe organization; if an individual is the cause of the mistreatment,then deviance will most likely be directed against the individual.Therefore, we theorize that centralization may be positively relatedto DWBI and DWBO.

H3a. Centralization is positively related to DWBI.

H3b. Centralization is positively related to DWBO.

On the other hand, Matheson (2007) discovered that centralizedorganizations will increase work alienation, which will in turnpromote employees workplace friendship (Sias & Cahill, 1998).Higher workplace friendship implies that employees are morewilling to assist other colleagues (Bowler & Brass, 2006), whichreinforces affective support among employees (Berman, West, &Richter, 2002). Therefore, employees would be less willing toexhibit DWB (Dalal, 2005). Workplace friendship refers to informaland personal-related interactions in a workplace setting (Bermanet al., 2002). Workplace friendship increases support andresources that help individuals do their jobs, which will in turnreduce DWB. According to the above, we theorize that higherorganizational centralization may be associated with loweremployee DWBI and DWBO.

According to the above literature review and since the effect ofcentralization on DWB is somewhat ambivalent, this researchproposes the following contrasting hypotheses:

H4a. Centralization is negatively related to DWBI.

H4b. Centralization is negatively related to DWBO.

2.7. Mediating effect of procedural justice

Procedural justice refers to the perceived fairness of the meansused to determine the outcomes employees receive (Folger &Konovsky, 1989). Since less participation allows less voice, andmore authority hierarchy allows less choice, then procedural justicedecreases with each (Schminke et al., 2002). Schminke, Ambrose,and Cropanzano (2000) also indicated that participation in deci-sion-making is positively related to procedural justice andauthority hierarchy is negatively related to procedural justice.Therefore, higher centralization is associatedwith lower proceduraljustice. Ambrose and Schminke (2003) suggested that underdifferent organizational structures, procedural justice will playdifferentially key roles in determining the quality of supervisorysocial exchange and organizational social exchange. Organizationalare receiving unfavorable treatment are more likely to feel angry,vengeful, and dissatised. Employees may retaliate against dissat-isfying conditions and unjust workplaces by engaging in behaviorthat harms the organization or other employees. Research sug-gested that dissatised employees often resort to deviant behaviorsas a way of coping with frustration (Judge et al., 2006). Rusbult,Farrell, Rogers, and Mainous (1988) also stated that employeesmay express their dissatisfaction of work through deviant behav-iors. Hence, Mount, Ilies, and Johnson (2006) asserted that thoseemployees who are less satised with their jobs are more likely todisplay both interpersonal and organizational DWB. Accordingly,higher job dissatisfaction could have an impact on increasing thepotential for DWB (Bowling, 2010; Dalal, 2005; Judge et al., 2006).Many researchers have emphasized the importance of dis-

agement 36 (2013) 401e410 403justice facilitates the formation of social exchange relationships and

2007). In turn, these attitudes lead to DWB against the organiza-tion (Skarlicki & Folger, 1997). Therefore, we can theorize that the

Marelationship between centralization and OCB/DWB will be medi-ated by procedural justice.

H5a. Procedural justice will mediate the relationship betweencentralization and OCBI.

H5b. Procedural justice will mediate the relationship betweencentralization and OCBO.

H6a. Procedural justice will mediate the relationship betweencentralization and DWBI.

H6b. Procedural justice will mediate the relationship betweencentralization and DWBO.

3. Research method

3.1. Sample and data collection

The research hypotheses are examined by collecting data fromemployees in Taiwans international tourist hotels. Prior to datacollection, we pre-tested our questionnaire with a sample of thirtyemployees from an international tourist hotel. Based on thepretests, the active voice of three items (items 1, 5, 6 of proceduraljustice) was replaced by the passive voice in Chinese. The purposeof rephrasing statements was tomake the statements clearer and toexplicitly measure the construct. In fact, the statements areconsistent with the original statements in English (Niehoff &Moorman, 1993). Reliability of the measurements was estimatedusing Cronbachs coefcient alpha, and values were well above thesuggested cutoff of 0.70 (Nunnally, 1978).

According to statistics obtained from the Tourism Bureau inTaiwan, that city had 69 international tourist hotels in 2010. Toinitiate a sample, this research contacted all human resourcemanagers of these hotels. Twenty-four hotels agreed to participateby providing a complete mailing list of employees names andaddresses. The size of these hotels ranged from 200 rooms to 700rooms (16 hotels have 200e400 rooms and 8 hotels have 400e700rooms). In January 2011, the author mailed questionnaires to 600thus fosters OCB (Moorman & Byrne, 2005). When employeesperceive the process by which outcome allocation decisions aremade to be fair, they are more likely to display OCBs (Niehoff &Moorman, 1993). Employees who perceive that they are treatedfairly by the organization may develop a sense of obligation toreciprocate by performing OCB (Lee & Allen, 2002; Zhang &Agarwal, 2009). Hence, a variety of studies have found a positiverelationship between perception of procedural justice and OCB(Nadiri & Tanova, 2010; Zhang & Agarwal, 2009).

Previous research also suggested that fairness perceptions playan important role in provoking DWB (Browning, 2008; Robinson &Bennett, 1995). If employees perceive managerial actions andorganizational decisions as unfair or unjust, they are apt to expe-rience feelings of resentment, anger, and outrage (Greenberg, 1990;Skarlicki, Folger, & Tesluk, 1999). Moreover, Hollinger and Clark(1983) found that if employees perceive unjust and unfair feel-ings, they would engage in actions in violation of the organizationto restore their unjust perceptions, such as stealing, going slow inwork, being late or leaving early, and absence. In addition, whenemployees perceive their organization as using unfair proceduresfor resource allocations, they will develop negative attitudestoward the organization (e.g., reduced trust and commitment andincreased dissatisfaction) (Zoghbi-Manrique & Verano-Tacoronte,

C.-H. Yen, H.-Y. Teng / Tourism404employees working in these hotels (average of 20e30 employeesper hotel). Of all the samples, 24 different hotels responded byreturning at least 10 questionnaires. A total of 368 questionnaireswere received, representing a response rate of 61.33 percent.Among the returned questionnaires, 26 incomplete surveys wereremoved due to missing data, and 24 respondents had worked fortheir current organization for less than 6months. As a result, usablequestionnaires totaled 318.

3.2. Measures

The centralization instrument consisted of ve items (Ferrell &Skinner, 1988). Employees indicated their agreement with eachitem, using a ve-point Likert scale ranging from StronglyDisagree to Strongly Agree.

We measured OCB using a 14-item, ve-point Likert scale,ranging from Strongly Disagree to Strongly Agree (Williams &Anderson, 1991). In this study, two types of employee behaviorswere measured, including display of OCBs that immediatelybeneted specic individuals and indirectly beneted the organi-zation (OCBI), along with those behaviors that beneted the orga-nization as a whole (OCBO). Seven items each measured the OCBdimensions of OCBI and OCBO. Higher scores reected higher levelsof OCB.

Wemeasured DWB using a 19-item (Bennett & Robinson, 2000),ve-point scale, ranging from Never to Daily according toDunlop and Lee (2004) andMount et al. (2006). This DWB scale wasexpected to make a distinction between deviant behavior directedagainst the organization (DWBO: 12 items) and that against indi-viduals at work (DWBI: 7 items). Higher scores reected higherlevels of DWB.

Procedural justice was measured with a 6-item subscaledeveloped by Niehoff and Moorman (1993) and a ve-point Likertscale, ranging from Strongly Disagree to Strongly Agree.

Questions relating to demographic data, included gender, age,education, and organizational tenure, were also included in thequestionnaire. The questionnaire was originally written in Englishand then translated into Chinese. Translationwas completed by theresearchers along with two other native English speakers whoworked in the hospitality industry. Before nalizing the question-naire design, back translation was done to reduce translation bias,as suggested by Van de Vijver and Hambleton (1996).

3.3. Data analysis

Data analyses for this study included descriptive analyses,conrmatory factor analysis (CFA), and multiple regression anal-ysis. The collected data were analyzed using the Statistical Packagefor Social Science (SPSS). Descriptive statistics included the mean,standard deviation of centralization, OCBs, DWBs and proceduraljustice. CFAwas used to assess the validity of the measures throughAMOS 7.0. This study tested the hypothesized relationshipsutilizing a series of regression analyses. The mediation tests fol-lowed the approach of Baron and Kenny (1986) and Sobel (1982).

4. Results

4.1. Characteristics of the sample

Among the 318 respondents,136 respondents (42.8%) weremaleand 182 respondents (57.2%) were female. 42.4% of the respondentswere aged between 21 and 30 years. Educational levels were fairlyhigh, with over 90% having college experience or above. Themajority (51.9%) of the respondents had average tenure of less than

nagement 36 (2013) 401e4105 years.

interrater agreement. Across the 24 hotels that constituted oursample, the IRC for centralization ranged from 0.74 to 0.92, sug-

Table 1Results of conrmatory factor analysis.

Constructs Factorloadings

Compositereliability

AVE

Centralization 0.83 0.50Any major decision that I makehas to have this companys approval

0.69

In my experience with this company,even quite small matters have to bereferred to someone higher up fora nal answer

0.77

My experiences with this companyare subject to a lot of rules andprocedures stating how variousaspects of my job are to be done

0.75

I have to ask senior managementbefore I do almost anything inmy business

0.73

I can take very little action onmy own until senior managementapprove it

0.59

OCBI 0.88 0.51Helps others who have been absents 0.56Helps others who haveheavy workloads

0.75

Assists supervisor with his/herwork (when not asked)

0.67

Takes time to listen to coworkersproblems and worries

0.73

Goes out of way to help new employees 0.58Takes a personal interest inother employees

0.86

Passes along information to co-workers 0.79OCBO 0.87 0.50Attendance at work is above the norm 0.74Gives advance notice when unableto come to work

0.81

Takes undeserved work breaks (R) 0.65Great deal of time spent with personalphone conversations (R)

0.78

Complains about insignicant thingsat work (R)

0.64

Conserves and protectsorganizational property

0.73

Adheres to informal rules devisedto maintain order

0.57

DWBI 0.90 0.58Made fun of someone at work 0.70Said something hurtful tosomeone at work

0.83

Made an ethnic, religious, orracial remark at work

0.74

Cursed at someone at work 0.75Played a mean prank on someone at work 0.68Acted rudely toward someone at work 0.82Publicly embarrassed someone at work 0.78

DWBO 0.92 0.51Taken property from workwithout permission

0.74

Spent too much time fantasizing ordaydreaming instead of working

0.69

Falsied a receipt to get reimbursedfor more money than you spenton business expenses

0.70

Taken an additional or longer breakthan is acceptable at your workplace

0.66

Come in late to work without permission 0.75Littered your work environment 0.73Neglected to follow your bosss instructions 0.71Intentionally worked slower thanyou could have worked

0.68

Discussed condential companyinformation with an unauthorized person

0.65

Used an illegal drug or consumedalcohol on the job

0.67

Put little effort into your work 0.75Dragged out work in order to get overtime 0.79

C.-H. Yen, H.-Y. Teng / Tourism Mangesting that aggregation was appropriate (George, 1990).

4.3. Measurement properties

To validate the constructs, the research model was estimatedwith the CFA in which all measurement items were loaded on theirexpected constructs, and the constructs were correlated in theanalysis (Gerbing & Anderson, 1988). Since the chi-square statisticis sensitive to sample size, this study relied on other indices in thetesting models (Hair, Black, Babin, Anderson, & Tatham, 2006). Inthe testing model for CFA, all factor loadings were signicant4.2. Aggregation of individual level data

Data for all the measures were obtained at the individual level.However, centralization reects an organizational level variable. Inorder to make the level of the data consistent with the level of thetheory, we aggregated data for this measure to the organizationallevel. We tested for interrater agreement using the interrateragreement coefcient (IRC) developed by James, Demaree, andWolf (1984). They recommended an IRC of 0.70 as an indicator of

Table 1 (continued )

Constructs Factorloadings

Compositereliability

AVE

Procedural justice 0.88 0.55Job decisions are made by the generalmanager in an unbiased matter

0.74

My general manager makes sure that allemployee concerns are heard before jobdecisions are made

0.81

To make job decisions, my general managercollects accurate and complete information

0.72

My general manager claried decisions andprovides additional information whenrequested by employees

0.70

All job decisions are applied consistentlyacross all affected employees

0.79

Employees are allowed to challenge orappeal job decisions made by thegeneral manager

0.77

Note: R refers to reversed question items; AVE refers to average variance extracted.

agement 36 (2013) 401e410 405(p < 0.001). The indexes of the model provide a good t:c2 1960.83, df 887, c2/df 2.21, GFI (goodness-of-tindex) 0.92, AGFI (adjusted goodness-of-t index) 0.90, RMSEA(root mean square error of approximation) 0.05, NFI (normed tindex) 0.91, and CFI (comparative t index) 0.95, which wasabove the model adaptability standard suggested by Hair et al.(2006) (c2/df < 3, GFI 0.90, AGFI 0.90, RMSEA 0.05,NFI 0.90, CFI 0.90), showing unidimensionality of the scales.Table 1 shows that the composite reliability ranged from 0.83 to0.92, greater than the standard of 0.6 (Hair et al., 2006). Theresearchers also employed a set of established procedures to checkfor convergent validity and discriminant validity of our scales. Theaverage variance extracted (AVE) for each construct was between0.50 and 0.58, which was either equal to or higher than 0.5 (Bagozzi& Yi, 1988; Fornell & Larcker, 1981), supporting convergent validity.The researchers measured discriminant validity by calculatingthe AVE for all pairs of constructs and comparing this value to thesquared correlation between the two constructs of interest. Theresearch results show that the squared correlation between anypair of constructs in all cases was less than the respective AVE ofeach of the constructs in the pair (Fornell & Larcker, 1981), sup-porting discriminant validity.

4.4. Common method bias checking

Common method bias may affect the empirical results becausethe data of this study were collected through self-report ques-tionnaires. Padsakoff, MacKenzie, Lee, and Podsakoff (2003) havedemonstrated that procedural and statistical techniques can beused to test for common method biases. In the procedural tech-nique, the researchers guaranteed respondents condentiality andanonymity to diminish the social desirability bias. In addition, theresearchers separated the items of centralization from those of

OCBO (b 0.28, p < .01). Thus, H2a and H2b were supported.Furthermore, both models with DWBI (F 23.98, p < .01) andDWBO (F 39.13, p < .01) as dependent variable were found to bestatistically signicant. Centralization was negatively related toDWBI (b0.27, p< .01) and DWBO (b0.34, p< .01). Thus, H4aand H4b were supported. However, the difference between thecoefcients was not statistically signicant (Paternoster, Brame,Mazerolle, & Piquero, 1998).

It was hypothesized that procedural justice would mediate therelationship between centralization and OCBI/OCBO and DWBI/DWBO. This assumption was tested with mediation analysis (Baron& Kenny, 1986). The rst step of mediation analysis is to show thatthe independent variable (centralization) affects the dependentvariables (OCBI/OCBO and DWBI/DWBO), as shown in Table 3. The

2 3 4 5 6

0.32**0.30** 0.45**

0.28** 0.37** 0.36**0.25** 0.31** 0.32** 0.64**

Table 3Regression results of centralization on OCB, DWB and procedural justice.

Independent variable OCBI OCBO DWBI DWBO Proceduraljustice

b b b b b

Centralization 0.33** 0.28** 0.27** 0.34** 0.19**R2 0.12 0.09 0.08 0.13 0.05Adjusted R2 0.10 0.08 0.07 0.11 0.04F 33.52** 24.76** 23.98** 39.13** 10.49**

*p < 0.05; **p < 0.01.

C.-H. Yen, H.-Y. Teng / Tourism Management 36 (2013) 401e410406employees perceived OCBs/DWBs, i.e., these two sections of theitems appeared on different pages of the questionnaire. This yiel-ded an effect of psychological separation on the respondents(Padsakoff et al., 2003).

In the statistical technique, thepossibility of commonmethodbiaswas tested using Harmans one-factor test (Podsakoff & Organ,1986).A principal component factor analysis was conducted on the items ofcentralization, OCBs/DWBs, and procedural justice. This yielded sixfactors with eigen values greater than 1. Cumulative loading was76.42% and factor loading of the rst factor was 20.66%. Also, theresearchers employed CFA to test the t of a one-factor model anda six-factor model. The results showed that the ts of the one-factormodel (where all items were loaded on a single factor) were worse(c2 3795.32, df 902, c2/df 4.21, GFI 0.74, AGFI 0.70,RMSEA 0.11, NFI 0.72, CFI 0.78).Moreover,we controlled for theeffects of a single unmeasured latent method factor in our analyses.We compared themeasurementmodelwithout the commonmethodvariance (CMV) factor andwith the factor. The t indices of themodelwith the CMV factor indicate thatc21920.61, df 878,c2/df 2.19,GFI 0.91, AGFI 0.89, RMSEA 0.05, NFI 0.90, CFI 0.94. Acomparison of the twomodels indicates that the change in t indiceswas not signicant. The factor loadings of these items remainedsignicant. The results suggest CMVwas not a major problem for thedata (Padsakoff et al., 2003).

4.5. Correlation analysis

Table 2 presents themeans, standard deviations, and correlationsof all variables, which provide an initial examination of the proposedrelationships. Centralization was positively related to OCBI (r 0.33,p< .01) and OCBO (r 0.28, p< .01), and negatively related to DWBI(r0.27, p< .01) andDWBO (r0.34, p< .01). Centralizationwasnegatively related to procedural justice (r 0.19, p < .01). Finally,procedural justicewaspositively related toOCBI (r0.32,p< .01)andOCBO (r 0.30, p < .01), and negatively related to DWBI (r 0.28,p < .01) and DWBO (r 0.25, p < .01).

4.6. Multiple regression analysis

To test the hypotheses more adequately, a series of regressionanalyses was conducted. All hypothesized independent variableswere entered into the equations for each dependent variable. Inaddition, we followed the three-step procedure described by Baron

Table 2Means, standard deviations and correlations of all variables.

Variable Mean SD 1

1.Centralization 3.58 0.612. Procedural Justice 3.45 0.72 0.19**3.OCBI 3.79 0.59 0.33**4.OCBO 3.88 0.56 0.28**5.DWBI 1.93 0.52 0.27**6.DWBO 1.82 0.57 0.34**N 318; *p < 0.05; **p < 0.01.and Kenny (1986) to test the mediating relationships. The rst stepwas to test the effect of independent variable on the dependentvariables. The second step was to test the effect of the independentvariable on the mediator variable. The third step was to test theeffects of both the mediator and independent variable on thedependent variables. If signicant relationships emerge in all threeregressions, then a partial mediation exists. If the independentvariable in the third step is not signicant, then a full mediationexists. The results of each step are shown in Table 3 and Table 4.

The relationships between centralization and OCBI/OCBO andDWBI/DWBO were tested in four regression models that includedcentralization as the independent variable (see Table 3). Bothmodels with OCBI (F 33.52, p< .01) and OCBO (F 24.76, p< .01)as the dependent variable were found to be statistically signicant.Centralization was positively related to OCBI (b 0.33, p < .01) and

Table 4Test of the mediating effect of procedural justice.

Independent variables OCBI OCBO DWBI DWBO

b b b b

Centralization 0.18* 0.15* 0.16* 0.19*Procedural justice 0.27** 0.26** 0.25** 0.22**R2 0.17 0.13 0.14 0.15Adjusted R2 0.16 0.12 0.13 0.14F 28.84** 20.31** 20.19** 21.76**

*p < 0.05; **p < 0.01.

independent variable. This model with procedural justice as the

st re

12159505

Mandependent variable was found to be statistically signicant(F 10.49, p < .01). Centralization was negatively related toprocedural justice (b 0.19, p < .01).

The nal step is to show that the mediator (procedural justice)affects the dependent variables (OCBI/OCBO and DWBI/DWBO)when the independent variable (centralization) is included in theequation. If procedural justice has a mediator effect, a signicantrelationship between centralization and OCBI/OCBO and DWBI/DWBO should disappear or be reduced when procedural justice isadded to the model. This analysis is shown in Table 4.

Four models that included procedural justice as the indepen-dent variable as well as centralization were found to be statisticallysignicant. In both models with OCBI and OCBO as the dependentvariables, the effects of centralization were reduced to b 0.18 andb 0.15 respectively, whereas procedural justice was positivelyrelated to OCBI (b 0.27, p < .01) and OCBO (b 0.26, p < .01). Inboth models with DWBI and DWBO as the dependent variables, theeffects of centralization were reduced to b 0.16 and b 0.19respectively, whereas procedural justice was negatively related toDWBI (b 0.25, p < .01) and DWBO (b 0.22, p < .01).

In addition, the Sobel test and a bootstrapping method wereused to examine the signicance of the mediating roles of proce-dural justice (Preacher, Rucker, & Hayes, 2007). The Sobel testshows whether the indirect effect of centralization on OCBs/DWBsvia procedural justice is different from zero. If a z score is largerthan 1.96, then the hypotheses regarding the indirect effect issubstantiated. The SPSS-macro provides an estimate of the trueindirect effect and its bias-corrected 95% condence interval (CI). InTable 5, both the Sobel test and the bootstrapping methoddemonstrated that the mediating roles of procedural justice on therelationship between centralization and OCBs/DWBs were signi-cant (OCBI: 95% CI 0.0733, 0.0294, p < .01; OCBO: 95%CI 0.0719, 0.0269, p < .01; DWBI: 95% CI 0.0289, 0.0661,p < .01; DWBO: 95% CI 0.0212, 0.0624, p < .01). Therefore, H5a,H5b, H6a and H6b were supported.

5. Conclusion and discussion

The purpose of this research is to examine the effect ofcentralization on OCB and DWB in the hospitality industry. Pastsecond step is to show that the independent variable (centraliza-tion) affects the mediator (procedural justice). In the second step,the relationship between centralization and procedural justice wastested in a regression model that included centralization as the

Table 5Indirect effects of centralization on OCB/DWB through procedural justice (Sobels te

Indirect effect Value Se

Centralization/procedural justice/OCBI 0.0513 0.01Centralization/procedural justice/OCBO 0.0494 0.01Centralization/procedural justice/DWBI 0.0475 0.00Centralization/procedural justice/DWBO 0.0418 0.01

C.-H. Yen, H.-Y. Teng / Tourismliterature offered inconsistent views on the relationships betweencentralization, OCB and DWB. According to literature, this studyestablished contrasting hypotheses. Empirical results demon-strated that centralization signicantly and positively affects OCBI/OCBO and signicantly and negatively relates to DWBI/DWBO.Findings indicated that from the standpoint of hospitalityemployees, higher degree of organizational centralization is asso-ciated with higher employee OCB and lower employee DWB.Moreover, procedural justice partially mediated the relationshipbetween centralization, OCBI/OCBO and DWBI/DWBO.The results of this study indicated that centralization is associ-ated with higher OCBI and OCBO. Centralization elicits employeeswork disinterest and lowers work enthusiasm (Matheson, 2007).Employees would transfer their focus to other activities, such asconversation (Matheson, 2007), which will in turn enhance work-place friendship (Sias & Cahill, 1998) and OCBI/OCBO. In addition,the results of this study suggested that centralization is stronglyrelated to OCBI. Raub (2008) suggested that centralization repre-sents a personal form of control which is expressed through theway in which managers or supervisors interact with their subor-dinates. Consequently, it could be that personalized controlmechanisms have a stronger inuence on employees.

The results of this study indicated that centralization is associ-ated with lower DWBI and DWBO. Consistent with the abovediscussion, higher workplace friendship will result in lower DWB.Workplace friendship increases support and resources from othersthat help individuals get their jobs done, which will in turn reduceDWBI. Organizations benet from supportive climates, which willin turn be linked to decreased DWBO.

From the standpoint of national culture, Oriental societiesemphasize collectivism (Hofstede, 1980). Although a higher degreeof centralization causes work inexibility, organizational normsand policies facilitate the smooth completion of objectives underthe inuence of collectivism. In other words, employees of Orientalsocieties still view overall performance objectives as the rstpriority. Also, collectivism is characterized as being tight, in whichgroup benets and in-group harmony are highly emphasized(Triandis, 1995). There is a consensus among members on whatappropriate behavior is in a particular situation (Pelto, 1968).Relatively speaking, employees of Western societies focus onindividualism. As a result, a higher degree of centralization wouldinhibit and negatively affect employees OCBs.

In addition, this research found that procedural justice is a keyfactor affecting employee behavior. A higher degree of proceduraljustice reinforces employee OCB and lowers DWB intentions. Thesendings show that procedural justice partially mediated the rela-tionship between centralization and employee discretionarybehavior. The above explains the possible reasons for the incon-sistent results of past studies and this research.

The results of this studymirror the relevant research reported inthe literature, such as Matheson (2007). A more centralized orga-nization enhances workplace friendship and reinforces affectiveties. It also raises employees OCB and lowers DWB intention.However, empirical ndings of this study contradict those ofStamper and Van Dyne (2003) and Raub (2008), which found thatcentralization negatively relates to OCB. The result of this research

sults).

LL 95% CI UL 95%CI Z Sig(two)

0.0733 0.0294 4.55 0.00000.0719 0.0269 4.41 0.00010.0289 0.0661 4.85 0.00000.0212 0.0624 4.00 0.0005

agement 36 (2013) 401e410 407is also inconsistent with the social exchange theory (Wat & Shaffer,2005). However, this nding is consistent with previous empiricalresearches (e.g., Jogaratnam & Tse, 2006). Within the Asian context,it would seem that decentralized structures tend to have a negativeeffect on performance, while centralized structures have a positiveeffect on performance. The result indicates that cultural valueswould likely override values in organizational culture (Laurent,1986), and this may help partially explain the positive associationbetween centralization and OCB, and the negative associationbetween centralization and DWB.

lessen the unnecessary misunderstanding and mission delays

visory support can enhance employees OCB intention (Chen &

MaChiu, 2008) and lower organizational deviance (Ferris, Brown, &Heller, 2009), and thus suggested offering employees more posi-tive encouragement and assistance, raise employees affectivedependence on the organization and the supervisor so thatemployees are willing to provide the organization with moreassistance and lessen the likelihood of DWB.

This research has several limitations. First, the cross-sectionaldesign limits the extent to which causeeeffect relations can beinferred from our research ndings. Second, DWB was measuredvia self-report. Respondents may fake good under the effect ofsocial desirability bias. Third, we did not consider the effect ofmanagement style (e.g., abusive or authoritarian managementstyle) on OCB and DWB. This factor might be considered as a limi-tation of this study and recommendations for future research.Fourth, this study measured OCB/DWB with 14/19-items andDWBO/DWBI with 12/7 items. The imbalance of data may causea bias in the outcome of our analyses. Finally, samples for the mainsurveywere drawn from the hospitality industry, whichmight limitthe generalizability of the results to other industries.

From an academic viewpoint, several areas for further researcharise from this study. First, we suggest future studies to makea comparison between industries and conduct more in-depthdiscussions on the effect of centralization on OCB and DWB.Second, future research that adopts a longitudinal design would bebetter suited to addressing the causal status of the variablesexamined in this research. Finally, empirical ndings of this studydiffer with those of Stamper and Van Dyne (2003) and Raub (2008).This is perhaps due to Orientals different work notions thanWestern employees. We thus suggest future studies to compareemployees of Oriental and Western countries and further examinethe impact of centralization on employee discretionary behavior.

Appendix A. Supplementary data

Supplementary data related to this article can be found at http://dx.doi.org/10.1016/j.tourman.2012.10.003.

References

Ambrose, M. L., & Schminke, M. (2003). Organization structure as a moderator ofthe relationship between procedural justice, interactional justice, perceivedorganizational support, and supervisory trust. Journal of Applied Psychology,88(2), 295e305.

Avey, J. B., Luthans, F., & Youssef, C. M. (2010). The additive value of positivepsychological capital in predicting work attitudes and behaviors. Journal ofManagement, 36(2), 430e452.

Bagozzi, R. P., & Yi, Y. (1988). On the evaluation of structural equation models.Journal of the Academy of Marketing Science, 16(1), 76e94.

Baron, R. M., & Kenny, D. A. (1986). The moderator-mediator variable distinction incaused by disparity of opinions. Also, organizations are recom-mended to strengthen employees identication and cohesion inthe organization, as well as elevate employee OCB and reduce DWB.Third, past studies showed that organizational support and super-Results of this study possess several managerial implications.First, although previous researchers found that a centralizedorganization is negatively related to OCBs (Raub, 2008), and posi-tively related to DWBs, the ndings of this research suggest thatWestern theories cannot be easily generalized to Oriental countries.This study strongly indicates the importance of centralized orga-nizations for individuals in the Oriental context. Second, werecommend that hospitality managers strengthen the verticalcommunication channels. Through direct contact, hospitalityorganizations can reduce the time required to executemissions and

C.-H. Yen, H.-Y. Teng / Tourism408social psychological research: conceptual strategic and statistical consider-ations. Journal of Personality and Social Psychology, 51(6), 1173e1182.Bennett, R. J., & Robinson, S. L. (2000). Development of a measure of workplacedeviance. Journal of Applied Psychology, 85(3), 349e360.

Berman, E. M., West, J. P., & Richter, M. N. (2002). Workplace relations: friendshippatterns and consequences (according to managers). Public AdministrationReview, 62(2), 217e230.

Blau, P. M. (1964). Exchange and power in social life. New York: Wiley.Bowler, W. M., & Brass, D. J. (2006). Relational correlates of interpersonal citi-

zenship behavior: a social network perspective. Journal of Applied Psychology,91(1), 70e82.

Bowling, N. (2010). Effects of job satisfaction and conscientiousness on extra rolebehaviors. Journal of Business and Psychology, 25(1), 119e130.

Browning, V. (2008). An exploratory study into deviant behaviour in the serviceencounter: how and why front-line employees engage in deviant behaviour.Journal of Management and Organization, 14(4), 451e471.

Chen, C. C., & Chiu, S. F. (2008). An integrative model linking supervisor support andorganizational citizenship behavior. Journal of Business and Psychology, 23(1/2),1e10.

Cohen, A., & Avrahami, A. (2006). The relationship between individualism, collec-tivism, the perception of justice, demographic characteristics and organisa-tional citizenship behaviour. The Service Industries Journal, 26(8), 889e901.

Colbert, A. E., Mount, M. K., Harter, J. K., Witt, L. A., & Barrick, M. R. (2004). Inter-active effects of personality and perceptions of the work situation on workplacedeviance. Journal of Applied Psychology, 89(4), 599e609.

Colquitt, J. A., Conlon, D. E., Wesson, M. J., Porter, C. O. L. H., & Ng, K. Y. (2001). Justiceat the millennium: a meta-analytic review of 25 years of organizational justiceresearch. Journal of Applied Psychology, 86(3), 425e445.

Conger, J. A., & Kanungo, R. N. (1988). The empowerment process: integratingtheory and practice. Academy of Management Review, 13(3), 471e482.

Dalal, R. S. (2005). A meta-analysis of the relationship between organizationalcitizenship behavior and counterproductive work behavior. Journal of AppliedPsychology, 90(6), 1241e1255.

DeGroot, T., & Brownlee, A. L. (2006). Effect of department structure on the orga-nizational citizenship behavior-department effectiveness relationship. Journalof Business Research, 59(10/11), 1116e1123.

Dineen, B. R., Lewicki, R. J., & Tomlinson, E. C. (2006). Supervisory guidance andbehavioral integrity: relationships with employee citizenship and deviantbehavior. Journal of Applied Psychology, 91(3), 622e635.

Dunlop, P., & Lee, K. (2004). Workplace deviance, organizational citizenshipbehavior, and business unit performance: the bad apples do spoil the wholebarrel. Journal of Organizational Behavior, 25(1), 67e80.

Erstad, M. (1997). Empowerment and organizational change. International Journal ofContemporary Hospitality Management, 9(7), 325e333.

Ertrk, A. (2007). Increasing organizational citizenship behaviors of Turkishacademicians: mediating role of trust in supervisor on the relationship betweenorganizational justice and citizenship behaviors. Journal of ManagerialPsychology, 22(3), 257e270.

Ferrell, O. C., & Skinner, S. J. (1988). Ethical behavior and bureaucratic structurein marketing research organizations. Journal of Marketing Research, 25(1),103e109.

Ferris, D. L., Brown, D. J., & Heller, D. (2009). Organizational supports and organi-zational deviance: the mediating role of organization-based self-esteem.Organizational Behavior and Human Decision Processes, 108(2), 279e286.

Fisher, R., McPhail, R., & Menghetti, G. (2010). Linking employee attitudes andbehaviors with business performance: a comparative analysis of hotels inMexico and China. International Journal of Hospitality Management, 29(3),397e404.

Folger, R., & Konovsky, M. A. (1989). Effects of procedural and distributive justiceon reactions to pay raise decisions. Academy of Management Journal, 32(1),115e130.

Fornell, C., & Larcker, D. (1981). Evaluating structure equations models withunobservable variables and measurement error. Journal of Marketing Research,18(1), 39e50.

George, J. (1990). Personality, affect, and behavior in groups. Journal of AppliedPsychology, 75(2), 107e116.

Gerbing, D. W., & Anderson, J. C. (1988). An updated paradigm for scale develop-ment incorporating unidimensionality and its assessment. Journal of MarketingResearch, 25(2), 186e192.

Gilbert, S., Laschinger, H. K. S., & Leiter, M. (2010). The mediating effect of burnouton the relationship between structural empowerment and organizational citi-zenship behaviours. Journal of Nursing Management, 18(3), 339e348.

Gold, A. H., Malhotra, A., & Segars, A. H. (2001). Knowledge management: anorganizational capabilities perspective. Journal of Management InformationSystems, 18(1), 185e214.

Greenberg, J. (1990). Organizational justice: yesterday, today, and tomorrow. Journalof Management, 16(2), 399e432.

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

Harper, D. (1990). Spotlight abuse-save prots. Industrial Distribution, 79(3),47e51.

Harris, L. C., & Ogbonna, E. (2006). Service sabotage: a study of antecedents andconsequences. Journal of the Academy of Marketing Science, 34(4), 543e558.

Hofstede, G. (1980). Motivation, leadership, and organization: do American theoriesapply abroad? Organization Dynamics, 9(1), 42e63.

nagement 36 (2013) 401e410Hofstede, G. (1991). Cultures and organizations: Software of the mind. New York:McGraw-Hill.

ManHollinger, R. C., & Clark, J. P. (1983). Deterrence in the workplace: perceivedcertainty, perceived severity, and employee theft. Social Forces, 62(2), 398e418.

Huang, X., Iun, J., Liu, A., & Gong, Y. (2010). Does participative leadership enhancework performance by inducing empowerment or trust? The differential effectson managerial and non-managerial subordinates. Journal of OrganizationalBehavior, 31(1), 122e143.

Jackson, C. L., Colquitt, J. A., Wesson, M. J., & Zapata-Phelan, C. P. (2006). Psycho-logical collectivism: a measurement validation and linkage to group memberperformance. Journal of Applied Psychology, 91(4), 884e899.

James, L. R., Demaree, R. G., & Wolf, G. (1984). Estimating within group interraterreliability with and without response bias. Journal of Applied Psychology, 69(1),85e98.

Jogaratnam, G., & Tse, E. C. Y. (2006). Entrepreneurial orientation and thestructuring of organizations: performance evidence from the Asian hotelindustry. International Journal of Contemporary Hospitality Management,18(6), 454e468.

Judge, T. A., Scott, B. A., & Ilies, R. (2006). Hostility, job attitudes, and workplacedeviance: test of a multi-level model. Journal of Applied Psychology, 91(1),126e138.

Kim, S., ONeill, J. W., & Cho, H. M. (2010). When does and employee not helpcoworkers? The effect of leader-member exchange on employee envy andorganizational citizenship behavior. International Journal of HospitalityManagement, 29(3), 530e537.

Lapierre, L. M., & Hackett, R. D. (2007). Trait conscientiousness, leader-memberexchange, job satisfaction and organizational citizenship behaviour: a test ofan integrative model. Journal of Occupational and Organizational Psychology,80(3), 539e554.

Laurent, R. (1986). The cross-cultural puzzle of international human resourcemanagement. Human Resource Management, 25(1), 91e102.

Lee, K., & Allen, N. J. (2002). Organizational citizenship behavior and workplacedeviance: the role of affect and cognitions. Journal of Applied Psychology, 87(1),131e142.

Lee, H. W., & Liu, C. H. (2007). An examination of factors affecting repatriatesturnover intentions. International Journal of Manpower, 28(2), 122e134.

Lin, L. H., & Ho, Y. L. (2010). Guanxi and OCB: the Chinese cases. Journal of BusinessEthics, 96(2), 285e298.

Markus, H. R., & Kitayama, S. (1991). Culture and self: implications for cognition,emotion, and motivation. Psychological Review, 98(2), 224e253.

Matheson, C. (2007). In praise of bureaucracy? A dissent from Australia. Adminis-tration & Society, 39(2), 233e261.

McGurn, J. (1988). Spotting the thieves who work among us. Wall Street Journal, 7March, 164.

Moorman, R. H., & Byrne, Z. S. (2005). How does organizational justice affectorganizational citizenship behavior? In J. Greenberg, & J. A. Colquitt (Eds.),Handbook of organizational justice (pp. 355e380) Mahwah, NJ: Erlbaum.

Mount, M., Ilies, R., & Johnson, E. (2006). Relationship of personality traits andcounterproductive work behaviors: the mediating effects of job satisfaction.Personnel Psychology, 59(3), 591e622.

Mulki, J. P., Jaramillo, F., & Locander, W. B. (2006). Emotional exhaustion andorganizational deviance: can the right job and a leaders style make a differ-ence? Journal of Business Research, 59(3), 1222e1230.

Nadiri, H., & Tanova, C. (2010). An investigation of the role of justice in turnoverintentions, job satisfaction, and organizational citizenship behavior in hospi-tality industry. International Journal of Hospitality Management, 29(1), 33e41.

Niehoff, B. P., & Moorman, R. H. (1993). Justice as a mediator of the citizenshipbetween methods of monitoring and organizational citizenship behavior.Academy of Management Journal, 36(3), 527e556.

Nunnally, J. (1978). Psychometric theory (2nd ed).. New York: McGraw-Hill.gaard, T., Marnburg, E., & Larsen, S. (2008). Perceptions of organizational structure

in the hospitality industry: consequences for commitment, job satisfaction andperceived performance. Tourism Management, 29(4), 661e671.

Organ, D. W., Podsakoff, P. M., & MacKenzie, S. B. (2006). Organizational citizenshipbehavior: Its nature, antecedents, and consequences. London: Sage.

Paternoster, R., Brame, R., Mazerolle, P., & Piquero, A. (1998). Using the correctstatistical test for the equality of regression coefcients. Criminology, 36(4),859e866.

Pavia, N., & Pilepic, L. (2010) Organizational structure in bringing about qualitativechange to hotel offer Tourism & Hospitality Management 2010, ConferenceProceedings, 1123e1128

Pelto, P. J. (1968). The difference between tight and loose societies. Society, 5(5),37e40.

Pertusa-Ortega, E. M., Zaragoza-Sez, P., & Claver-Corts, E. (2010). Can formaliza-tion, complexity, and centralization inuence knowledge performance. Journalof Business Research, 63(3), 310e320.

Pittaway, L., Carmouche, R., & Chell, E. (1998). The way forward: leadership researchin the hospitality industry. International Journal of Hospitality Management,17(4), 407e426.

Podsakoff, P. M., MacKenzie, S. B., Lee, J. Y., & Podsakoff, N. P. (2003). Commonmethod biases in behavioral research: a critical review of the literature andrecommended remedies. Journal of Applied Psychology, 88(5), 879e903.

Podsakoff, P., & Organ, D. (1986). Self-reports in organizational research: problemsand prospects. Journal of Management, 12(4), 531e544.

Podsakoff, N. P., Whiting, S. W., Podsakoff, P. M., & Blume, B. D. (2009). Individual-

C.-H. Yen, H.-Y. Teng / Tourismand organizational-level consequences of organizational citizenship behaviors:a meta-analysis. Journal of Applied Psychology, 94(1), 122e141.Porter, L. W., Lawler, E. E., III, & Hackman, J. R. (1996). Ways groups inuenceindividual work effectiveness. In R. M. Steers, L. W. Porter, & G. A. Bigley (Eds.),Motivation and leadership at work (pp. 346e354). New York: McGraw-Hill.

Preacher, K. J., Rucker, D. D., & Hayes, A. F. (2007). Addressing moderated mediationhypotheses: theory, methods, and prescriptions. Multivariate BehavioralResearch, 42(1), 185e227.

Raub, S. (2008). Does bureaucracy kills individual initiative? The impact of structureon organizational citizenship behavior in the hospitality industry. InternationalJournal of Hospitality Management, 27(2), 179e186.

Robinson, S. L., & Bennett, R. J. (1995). A typology of deviant workplace behaviors:a multidimensional scaling study. Academy of Management Journal, 38(2),555e572.

Robinson, S. L., & Bennett, R. J. (1997). Workplace deviance: its denition, itsmanifestations, and its causes. In R. J. Lewicki, R. J. Bies, & B. H. Sheppard (Eds.),Research on negotiation in organizations (pp. 3e27). Greenwich, CT: JAI Press.

Rusbult, C. R., Farrell, D., Rogers, G., & Mainous, A. G. (1988). Impact of exchangevariables on exit, voice, loyalty and neglect: an integrative model of responsesto declining job satisfaction. Academy of Management Journal, 31(3), 599e627.

Schminke, M., Ambrose, M. L., & Cropanzano, R. (2000). The effect of organizationalstructure on perceptions of procedural fairness. Journal of Applied Psychology,85(2), 294e304.

Schminke, M., Cropanzano, R., & Rupp, D. E. (2002). Organizational structure andfairness perceptions: the moderating effect of organizational level. Organiza-tional Behavior and Human Decision Processes, 89(1), 881e895.

Shenkar, O., & Von Glinow, M. A. (1994). Paradoxes of organizational theory andresearch: using the case of China to illustrate national contingency. Manage-ment Science, 40(1), 56e71.

Sias, P. M., & Cahill, D. J. (1998). From coworkers to friends: the development ofpeer friendships in the workplace. Western Journal of Communication, 62(3),273e299.

Skarlicki, D. P., & Folger, R. (1997). Retaliation in the workplace: the roles ofdistributive, procedural, and interactional justice. Journal of Applied Psychology,82(3), 434e443.

Skarlicki, D. P., Folger, R., & Tesluk, P. (1999). Personality as a moderator in therelationship between fairness and retaliation. Academy of Management Journal,42(1), 100e108.

Sobel, M. E. (1982). Asymptotic condence intervals for indirect effects in structuralequations models. In S. Leinhart (Ed.), Sociological methodology (pp. 290e312).San Francisco: Jossey-Bass.

Stamper, C. L., & Van Dyne, L. (2003). Organizational citizenship: a comparisonbetween part-time and full-time service employee. Cornell Hotel and RestaurantAdministration Quarterly, 44(1), 33e42.

Tan, H. H., & Tan, M. L. (2008). Organizational citizenship behavior and socialloang: the role of personality, motives, and contextual factors. Journal ofPsychology, 142(1), 89e108.

Taormina, R. J., & Gao, J. H. (2010). A research model for guanxi behavior: ante-cedents, measures, and outcomes of Chinese social networking. Social ScienceResearch, 39(6), 1195e1212.

Tracey, J. B., & Hinkin, T. R. (1996). How transformational leaders lead in thehospitality industry. International Journal of Hospitality Management, 15(2),165e176.

Tracey, B. J., & Nathan, A. E. (2002). The strategic and operational roles of humanresources: an emerging model. Cornell Hotel and Restaurant AdministrationQuarterly, 43(4), 17e26.

Triandis, H. C. (1995). Individualism and collectivism. Boulder, CO: Westview Press.Van de Vijver, F., & Hambleton, R. (1996). Translating tests: some practical guide-

lines. European Psychologist, 1(2), 89e99.Van Dyne, L., Graham, J. W., & Dienesch, R. M. (1994). Organizational citizenship

behavior: construct redenition, measurement and validation. Academy ofManagement Journal, 37(4), 765e802.

Van Yperen, N. W., van den Berg, A. E., & Willering, M. C. (1999). Towards a betterunderstanding of the link between participation in decision-making andorganizational citizenship behaviour: a multilevel analysis. Journal of Occupa-tional and Organizational Psychology, 72(3), 377e392.

Wat, D., & Shaffer, M. A. (2005). Equity and relationship quality inuences onorganizational citizenship behaviors: the mediating role of trust in the super-visor and empowerment. Personnel Review, 34(4), 406e422.

Willem, A., Buelens, M., & Jonghe, I. D. (2007). Impact of organizational structure onnurses job satisfaction: a questionnaire survey. International Journal of NursingStudies, 44(6), 1011e1020.

Williams, L. J., & Anderson, S. E. (1991). Job satisfaction and organizationalcommitment as predictors of organizational citizenship and in-role behaviors.Journal of Management, 17(3), 601e617.

Wong, Y. H., & Tam, J. L. M. (2000). Mapping relationships in China: guanxi dynamicapproach. Journal of Business & Industrial Marketing, 15(1), 57e70.

Zhang, H., & Agarwal, N. C. (2009). The mediating roles of organizational justice onthe relationships between HR practices and workplace outcomes: an investi-gation in China. The International Journal of Human Resource Management, 20(3),676e693.

Zheng, W., Yang, B., & McLean, G. N. (2010). Linking organizational culture, struc-ture, strategy, and organizational effectiveness: mediating role of knowledgemanagement. Journal of Business Research, 63(7), 763e771.

Zoghbi-Manrique, P., & Verano-Tacoronte, D. (2007). Investigating the effects of

agement 36 (2013) 401e410 409procedural justice on workplace deviance. International Journal of Manpower,28(8), 715e729.

Chang-Hua Yen is an Associate Professor in Department ofLeisure and Recreation Management, National TaichungUniversity of Science and Technology, Taiwan. He earnedhis PhD from the Chinese Culture University in Taiwan. Hisareas of research include service marketing and manage-ment, and tourism management.

Hsiu-Yu Teng is a PhD Student in Graduate Institute ofRecreation, Tourism and Hospitality Management,National Chiayi University, Taiwan. Her areas of researchinclude tourism management, and travel marketing.

C.-H. Yen, H.-Y. Teng / Tourism Management 36 (2013) 401e410410

The effect of centralization on organizational citizenship behavior and deviant workplace behavior in the hospitality industry1. Introduction2. Literature review and research hypotheses2.1. Organizational structure in the hospitality industry2.2. Organizational citizenship behavior2.3. Deviant workplace behavior2.4. Cross-cultural differences2.5. Centralization and organizational citizenship behavior2.6. Centralization and deviant workplace behavior2.7. Mediating effect of procedural justice

3. Research method3.1. Sample and data collection3.2. Measures3.3. Data analysis

4. Results4.1. Characteristics of the sample4.2. Aggregation of individual level data4.3. Measurement properties4.4. Common method bias checking4.5. Correlation analysis4.6. Multiple regression analysis

5. Conclusion and discussionAppendix A. Supplementary dataReferences