Transcript
Page 1: HOW DOES BUREAUCRACY IMPACT INDIVIDUAL.pdf

HOW DOES BUREAUCRACY IMPACT INDIVIDUALCREATIVITY? A CROSS-LEVEL INVESTIGATION OF TEAM

CONTEXTUAL INFLUENCES ON GOALORIENTATION–CREATIVITY RELATIONSHIPS

GILES HIRSTMonash University

DAAN VAN KNIPPENBERGErasmus University Rotterdam

CHIN-HUI CHENTaiwan Customs Bureau

CLAUDIA A. SACRAMENTOAston University

Offering important counterpoint to work identifying team influences stimulating cre-ative expression of individual differences in goal orientation, we develop cross-leveltheory establishing that team bureaucratic practices (centralization and formalization)constrain creative expression. Speaking to the tension between bureaucracy and cre-ativity, findings indicate that this influence is not only negative and that effects ofcentralization and formalization differ. Surveying 330 employees in 95 teams at theTaiwan Customs Bureau, we found that learning and “performance avoid” goal ori-entations had, respectively, stronger positive and weaker negative relationships withcreativity under low centralization. A “performance-prove” orientation was positivelyrelated to creativity under low formalization.

As employee creativity is crucial for organization-al innovation and survival (Amabile, 1988; Oldham& Cummings, 1996), managers and scholars alikehave sought to identify the ingredients that fosterindividual creativity. It is well recognized that theteam context in which employees are embeddedplays a central role in stimulating the creative ex-pression of individual differences (Amabile &Conti, 1999; Hirst, van Knippenberg, & Zhou, 2009;Shalley, Zhou, & Oldham, 2004). Accordingly, re-searchers have begun to adopt a cross-level focus toexamine the interplay between individual andteam factors (Zhou & Shalley, 2008; cf. Klein &Kozlowski, 2000). In this respect, research hasidentified individual differences in goal orientation

that encourage self-regulation in achievement situ-ations as a powerful influence on creativity when acontext stimulates their expression (Hirst et al.,2009). Yet researchers have in effect turned a blindeye to the fact that organizations, and units withinthem, also need to impose practices and proceduresthat themselves regulate, order, and control behav-ior (Burns & Stalker, 1961; Thompson, 1965). Orga-nizations and organizational units instill such prac-tices to ensure consistency, efficiency, and control(Adler, 1999), but such bureaucracy may stifle in-dividuals’ creativity. To build toward a comprehen-sive understanding of the factors affecting individualcreativity in teams, scholars thus need to consider notonly contextual factors that invite creativity, but alsocontextual influences that may constrain it. To pro-vide this important complementary perspective onthe current state of the science, we zoom in on teambureaucratic context and how it influences the cre-ative expression of goal orientations.

We rely on an integrative person-in-situation the-ory that describes how situational influences mayeither restrain or invite the expression of individ-ual differences (Tett & Burnett, 2003; cf. Mischel,1977). Although at first glance it might appear that

The authors would like to thank Associate Editor Eliz-abeth Morrison and the three anonymous reviewers fortheir insightful comments and suggestions. We alsothank Jeremy Dawson, Pamela Tierney, and Adam Grantfor their advice and The Faculty of Business and Eco-nomics, University of Melbourne, for the support andfacilities provided during the first author’s sabbatical.

Editor’s Note: The manuscript for this article was ac-cepted during Duane Ireland’s term as editor.

� Academy of Management Journal2011, Vol. 54, No. 3, 624–641.

624

Copyright of the Academy of Management, all rights reserved. Contents may not be copied, emailed, posted to a listserv, or otherwise transmitted without the copyright holder’s expresswritten permission. Users may print, download or email articles for individual use only.

Page 2: HOW DOES BUREAUCRACY IMPACT INDIVIDUAL.pdf

the influence of bureaucracy is straightforward—itsuppresses creativity—we propose that a closerlook suggests that the issue is more complex. First,there are important individual differences in goalorientations that capture individuals’ likelihood toengage in, or avoid, creative behavior, and there isa strong case to be made that these differences ingoal orientations lead individuals to responddifferently to higher and lower levels of teambureaucracy. Second, bureaucracy can be charac-terized in terms of two core dimensions, central-ization and formalization (Bolin & Harenstam,2008; Caruana, Morris, & Vella, 1998; Raub,2007), and we propose that high versus low levelsof centralization have different influences on in-dividuals’ creative tendencies than high versuslow levels of formalization.

Our core contribution lies in an important ad-vancement of person-in-situation analyses of cre-ativity (Hirst et al., 2009; Taggar, 2002), demon-strating that contextual influences may not onlyinvite the creative expression of individual differ-ences, but also constrain them. Our study thus pro-vides insights into resolution of the tension be-tween bureaucratic control and innovation, whichhas eluded the field for half a century (cf. Burns &Stalker, 1961). This is also a pragmatically impor-tant issue, as organizations rely on a certain level ofbureaucracy, prioritizing establishing and stickingto a beaten track, while also desiring creativity—which by definition entails stepping off the beatentrack. In examining person-in-situation influences,we deviate from the organizational design litera-ture’s study of the main effects of bureaucracy (e.g.,Bunderson & Boumgarden, 2010; Raub, 2007) andbring together two perspectives on behavioral reg-ulation that have been studied in separate tradi-tions. One perspective has been characterized byindividual-level analyses of self-regulation in goal-directed behavior (DeShon & Gillespie, 2005;Porath & Bateman, 2006); the other, by analyses ofcontextual regulation of behavior at the level ofsocial aggregates (Raub, 2007).

THEORETICAL BACKGROUNDAND HYPOTHESES

Individual creativity at work involves the devel-opment of practical and new solutions to work-place challenges, providing a tangible and usefuloutcome for an organization (Amabile, 1988). Fol-lowing from the description of creativity as anoutcome that derives from addressing work chal-lenges, it is not surprising that individual differ-ences in goal orientation that relate to an individ-ual’s motivation to tackle challenging problems

influence employee creativity (Gong, Huang, & Farh,2009; Hirst et al., 2009). We first outline this individ-ual difference perspective on creativity and then in-troduce the bureaucratic team context perspective,before we move on to our cross-level integration ofthese two perspectives in a series of hypotheses.

Individual Team Members: Goal Orientationsand Creativity

Achievement motivation theory describes goalorientations as motivational orientations that cap-ture how individuals regulate attention and effortwhen approaching, interpreting, and responding toachievement situations (DeShon & Gillespie, 2005;Elliot & Church, 1997). Two distinct orientationsare commonly identified. A learning goal orientation(from here on, a learning orientation) is focused onthe development of competence and task mastery andfosters an intrinsic interest in a task itself (Dweck,1999). Intrinsic task motivation encourages individu-als to invest effort and show perseverance (Amabile,1996), and it is not surprising that a learning orienta-tion encourages people to develop creative solutionsto problems at work (Gong et al., 2009; Hirst et al.,2009; cf. Janssen & Van Yperen, 2004).

People may also be motivated by extrinsic factorssuch as competing against others, receiving re-wards, acknowledgement, or avoiding criticism(VandeWalle, 1997). This motivation is captured bythe performance goal orientation, which is focusedon the demonstration of competence to others. Thisexternally attuned motivation can be divided intotwo subdimensions. A performance-prove goal ori-entation (from here on, a prove orientation) encour-ages individuals to seek to attain favorable judg-ments, whereas people who are concerned aboutavoiding unfavorable competence judgments havea performance-avoid goal orientation (from hereon, an avoid orientation). The prove orientationmay dispose individuals to be more creative whencreativity is valued as a way to demonstrate com-petence (Hirst et al., 2009). The avoid orientation,in contrast, disposes individuals to be less creative,because creativity inherently holds a risk of failure,and the possibility of appearing incompetent dis-courages these individuals from engaging in riskyor challenging activities (VandeWalle, 1997) thatwould have provided opportunities for creativity.

Goal orientation is fundamentally about self-reg-ulation of behavior (Button, Mathieu, & Zajac,1996; DeShon & Gillespie, 2005). Goal orientationsencourage people to choose, either consciously orsubconsciously, to engage in certain types of behav-iors in achievement situations. For instance, insuch situations, individuals with high levels of

2011 625Hirst, van Knippenberg, Chen, and Sacramento

Page 3: HOW DOES BUREAUCRACY IMPACT INDIVIDUAL.pdf

learning orientation may choose to engage in adap-tive behaviors patterns such as selecting challeng-ing tasks, setting difficult goals, and persistingwhen obstacles are encountered. Therefore, to ad-equately model the behavioral outcomes of goalorientations, it is necessary to consider how theyinform responses to the context in which the be-havior is enacted (DeShon & Gillespie, 2005). Morebroadly, the strong emergence of person-in-situa-tion approaches (Chen & Kanfer, 2006; Tett & Bur-nett, 2003; cf. Kristoff, 1996) highlights that ananalysis of the interplay between individual andcontext is essential to predict the expression ofindividual dispositions. That is, the concept of dis-position should not be misrepresented to imply thatan individual will always behave in certain ways.Instead, a person-in-situation approach suggests mov-ing away from a “main effects” approach in which theinfluence of individual differences is assumed to beconstant. Rather, it implies an emphasis on the con-tingencies of disposition-outcome relationships.

Individual creativity is often enacted in the con-text of a team or work group (Taggar, 2002), andthis context may influence the relationship be-tween goal orientation and creativity. Hirst et al.(2009) studied this very issue, focusing on teamlearning behavior as a contextual influence stimu-lating the expression of goal orientations that areconducive to creativity. They demonstrated thatteam learning behavior helped “bring out” the pos-itive relationship between a learning orientationand creativity and between a prove orientation andcreativity. Further testifying to the viability of thiscontingency perspective, Hirst et al. (2009) did notfind relationships between goal orientations andcreativity across the board but rather, found thatthe relationships were contingent on team learningbehavior. These findings show that team dynamicsmay stimulate the expression of creative tenden-cies, yet they are mute on the issue that assumescenter stage in the current study: the possibilitythat team contextual influences may also constrainthe expression of creative tendencies.

The absence of creativity-stimulating team con-textual influences such as team learning behaviorin no way imposes constraints on individuals’ cre-ative behavior. Team bureaucratic practices, incontrast, represent a different class of team contex-tual influences in that they may impose exactlysuch constraints on creativity, and their influencecannot be extrapolated from earlier findings con-cerning creativity-stimulating influences (e.g., theabsence of bureaucracy is not necessarily stimulat-ing, just as the absence of creativity-simulating in-fluences is not necessarily restraining). Adding tothe complexity of the issue, the restraining influ-

ence of team bureaucratic practices may not havenegative creativity consequences across the board,as we argue in the following.

The Team Bureaucratic Context: Centralizationand Formalization

Teams, departments, and organizations differ inthe extent to which bureaucratic practices re-strain their members. Conceptual frameworks dis-tinguish two main dimensions in this respect(Bolin & Harenstam, 2008; Caruana et al., 1998; Raub,2007; cf. Burns & Stalker, 1961): centralization ofdecision making (Van de Ven & Ferry, 1980) andformalization of rules and procedures prescribingand controlling behavior (Hall, 1999). Both central-ized decision making and formal rules and proce-dures are ways of regulating and controlling em-ployee behavior—the essence of bureaucracy—andare associated with low employee discretion on thejob. Centralization relates to how power is distrib-uted in an organizational hierarchy and whetheremployees are encouraged to participate in deci-sion making (Hage & Aiken, 1967). Low centraliza-tion captures a context in which all employeesparticipate and are afforded discretion and oppor-tunities to act according to their own inclination. Ifdecisions must be referred up the chain of com-mand and made by a few superiors, centralizationis high. Formalization relates to the extent to whichrules are clearly specified and procedures stan-dardized. Increasing formalization reduces the ex-tent of employees’ freedom by prescribing proce-dures and potentially by sanctioning some coursesof action—providing specific directions as to ap-propriate actions, directing and enforcing these ac-tions, and constraining employees’ ability to en-gage in discretionary behaviors (Raub, 2007).

In the team context, centralization captures theextent to which within-team decision authority liessolely with a team’s leader (decision making iscentralized) or is shared between leader and mem-bers (decision making is decentralized and partici-pative). Low centralization thus reflects an activeinfluence on team members, who are expected toshare decision-making authority with their leader.In that sense, low centralization may in fact have aninfluence that is described as “empowering” inother literatures in which decentralized, participa-tive decision making is accorded an important rolein actively engaging and intrinsically motivatingemployees (Arnold, Arad, Rhoades, & Drasgow,2000; Chen, Kirkman, Kanfer, Allen, & Rosen, 2007;Kirkman & Rosen, 1997). Low team formalization,in contrast, merely reflects the absence of rules andprocedures regulating team member behavior and

626 JuneAcademy of Management Journal

Page 4: HOW DOES BUREAUCRACY IMPACT INDIVIDUAL.pdf

thus if anything is merely a “passive” influence onindividuals. Although high centralization and for-malization may thus be similarly restraining in reg-ulating and controlling team member behavior, wepropose that low levels of centralization and for-malization may in fact reflect markedly differentinfluences on employees—an issue that becomesapparent when one considers their cross-level in-teraction with goal orientations.

The importance of a focus on team bureaucracyas an influence on goal orientation-creativity rela-tionships is evident in a conceptual analysis byThompson (1965), who suggested that bureaucraticpractices demotivate and restrain engagement inentrepreneurial, nonroutine tasks and limit innova-tion (cf. Raub, 2007). Such an inhibiting influencedirectly relates to the issues at stake in the currentanalysis. Situational factors that constrain or pre-scribe behavior limit the scope for the expression ofindividual differences by creating “strong” situa-tions that override dispositions (Mischel, 1977). Incontrast, situational influences may also invite(“activate” [Tett & Burnett, 2003]) the expression ofindividual dispositions. We propose that team bu-reaucratic practices are particularly relevant in thisrespect, because the level of team bureaucracy mayboth constrain and invite the expression of individ-ual differences in goal orientations. In the follow-ing sections, we outline these propositions in moredetail for both centralization and formalization.

Individual Differences in a Team Context: GoalOrientation and Centralization

In teams, centralization relates to the extent towhich the leaders control and independently makedecisions concerning the teams as opposed to en-gendering participative decision making (Bolin &Harenstam, 2008). Centralized decision making re-duces opportunities for individuals to contributeoriginal thoughts or idiosyncratic novel views andto participate in discussions about important issuesfacing a team—the very issues that might benefitfrom individual creativity in meeting challengesand solving problems. In contrast, decentralizeddecision making encourages individuals to contrib-ute and participate in decisions and provides indi-viduals latitude to express their views (Arnold etal., 2000; Chen et al., 2007). In decentralized con-texts, team members share their views, which notonly promotes a greater variety of different views,but may also invite a more committed and proac-tive engagement with potential challenges facing ateam. The greater autonomy afforded by decentral-ization also fosters psychological ownership andempowerment and thus builds enthusiasm, com-

mitment to decisions, and intrinsic motivation, apowerful seed for creativity (Amabile, 1996). De-centralized practices promote a supportive climatethat conveys an invitation to voice and share one’sown perspective (Arnold et al., 2000) and thereforeare also likely to reassure and encourage a team’smembers to feel that their contributions are appre-ciated (Wagner, 1994). This context may stimulateemployees to grapple with less routine challengesand problems facing their team (Ahearne, Mathieu,& Rapp, 2005; Srivastava, Bartol, & Locke, 2006).Decentralized decision making thus may play animportant role in bringing out the tendencies toengage in the creative behavior captured by learn-ing and prove orientations and in attenuating ten-dencies to walk away from creative challenges as-sociated with an avoid orientation.

Learning-oriented people are motivated to learnand seek out tasks that are challenging becausethese provide greater opportunities for develop-ment. Centralization reduces opportunities to ex-plore and learn and diminishes employee’s control.Centralization thus attenuates or removes the cre-ative benefits of a learning orientation that derivefrom intrinsically motivated engagement with workchallenges. Put differently, centralization inhibits(cf. Tett & Burnett, 2003) the creative expression ofa learning orientation. Decentralized decision-mak-ing practices, in contrast, should appeal to learn-ing-oriented individuals, activating (Tett & Burnett,2003) their learning orientation, because of the op-portunity to explore and learn from different viewsand ideas and because of the challenge and learn-ing opportunities implicit in engaging with com-plex problems facing their team. This context mayeven create a positive cycle of interest and enthu-siasm, encouraging higher intrinsic motivation thatfurther fuels creativity (Amabile, 1996). Thus:

Hypothesis 1a. A learning orientation and cen-tralization interact to influence creativity: thelearning orientation is positively related to cre-ativity when centralization is low, but notwhen centralization is high.

Performance-oriented people seek to maximizerewards and minimize possible punishments usingenvironmental cues to decide which behaviors areappropriate (Seijts, Latham, Tasa, & Latham, 2004).They can be expected to focus on the cues, in thiscase the amount of centralization evident in a team’scontext, as a source of information for determiningthe choice of actions favored in the team. The natureof the influence of decentralized decision making onindividuals with prove orientations compared tothose with avoid orientations will differ, however.

2011 627Hirst, van Knippenberg, Chen, and Sacramento

Page 5: HOW DOES BUREAUCRACY IMPACT INDIVIDUAL.pdf

For prove-oriented people, motivation to engagein creative activities depends on the extent towhich they perceive they are likely to be recog-nized and rewarded for their behavior. Accord-ingly, we may expect individuals with a strongerprove orientation to be more responsive to the ex-tent to which their context encourages them tocontribute to team decisions and to provide sugges-tions to solve team problems. When prove-orientedindividuals work in a context where the leaderarrives at decisions with little consultation or invi-tation to participate, they may assume the leaderhas little interest in their suggestions or views andbecome less likely to engage in complex problemsolving and creativity to meet work challenges;centralization inhibits the creative expression of aprove orientation. In contrast, in decentralized con-texts, where team members are encouraged to con-tribute, prove-oriented people will construe teamdiscussions as a forum in which to demonstratecompetence. The chance to display their prowessmay tantalize them to display high levels of profi-ciency and to be acknowledged and recognized fortheir abilities. For prove-oriented people too, decen-tralization works as trait activating. We therefore ex-pect that a prove orientation is related to creativitywhen centralization is low rather than high.

Hypothesis 1b. A prove orientation and cen-tralization interact to influence creativity; theprove orientation is positively related to cre-ativity when centralization is low, but notwhen centralization is high.

For an avoid orientation and centralization, mat-ters are different. Unlike learning and prove orien-tations, which may inspire active engagement withjob challenges, avoid orientations predispose indi-viduals to stay clear of job challenges and problemsthat may invite acts of creativity. As a result, thelack of support and appreciation for employees’proactive engagement with the work that is con-veyed by high centralization may actually reinforceavoid-oriented individuals’ dispositional tenden-cies, whereas the active, engaging influence of de-centralized decision making may work to counteravoidant tendencies. For individuals with an avoidorientation, centralization creates a team climatewith limited support and few opportunities to pro-actively contribute to decisions and solve workchallenges that require creativity. In these settings,absence of explicit encouragement to creatively en-gage with important job issues may lead avoid-oriented people to perceive their leader as less thansupportive of experimental and potentially riskyactivities, reinforcing their tendency to avoid suchactivities, which would have provided greater oppor-

tunities for creativity. Thus, in contrast to its inhibit-ing influence on the relationships between learningand prove orientations and creativity, if anything cen-tralization may bolster the negative relationship be-tween an avoid orientation and creativity.

A different set of events occurs in decentralizedcontexts, where participative practices signal thatindividuals’ contributions and proactive engage-ment with work challenges are not only supportedand encouraged, but also appreciated and ex-pected. This setting fosters a climate in whichnorms support participating in decisions and en-gaging with work challenges. This climate maystimulate or at least reassure avoid-oriented indi-viduals that their proactive engagement with workchallenges is appreciated, even though these be-haviors may differ from their inherently cautiousinclinations. This sense should lower their dispo-sitional “barriers” to trying new approaches, mak-ing it easier for avoid-oriented individuals to clar-ify uncertainties by facilitating shared learningthrough team discussion and problem solving.Likewise, decentralized practices that foster inclu-sive decision making diminish the negative conse-quences of voicing an opinion that may be out-of-synch with the views of one’s team, thus loweringthe threshold for engaging in such behaviors, be-cause it signals both that false assumptions or er-rors are less likely to elicit negative feedback andthat team members are encouraged to discuss dif-ferent approaches. Thus, we predict:

Hypothesis 1c. An avoid orientation and cen-tralization interact to influence creativity; theavoid orientation is negatively related to cre-ativity when centralization is high, but notwhen centralization is low.

Individual Differences in Team Context: GoalOrientation and Team Formalization

Formalized team contexts are characterized byrules and standardized procedures that limit thechoice of behaviors and decisions an individualcan make. Rules and regulations effectively reduceindividuals’ opportunities to engage in discretion-ary behaviors and provide a setting that “presses”individuals to follow certain procedures and ap-proaches (Hall, 1999). Teams that are high in for-malization are characterized by clear behavioralprotocols involving administrative checks (as evi-denced by paperwork and administration) that reg-ulate and direct employees’ behavior. Thus, theyare prototypical instances of what Mischel (1977)called “strong situations.” As formalization in-creases, behavior is regulated and restrained to a

628 JuneAcademy of Management Journal

Page 6: HOW DOES BUREAUCRACY IMPACT INDIVIDUAL.pdf

greater degree and funneled and homogenized to asmall set of choices offering little freedom, reduc-ing the expression of individual differences. Interms of trait activation theory (Tett & Burnett,2003), high formalization inhibits the expression ofindividual dispositions. In less formalized teamcontexts, there will be fewer guidelines, more op-portunities for discretion, and thus more leeway forthe expression of individual differences. This is notto say, however, that the absence of formalizedpractices encourages creativity (i.e., unlike decen-tralized decision making, which plays a more ac-tive role in this respect). Rather, it points to greaterfreedom for the expression of individual differ-ences with fewer specific procedural directions tofollow, which may either stimulate or impair cre-ativity, according to an individual’s disposition.

Learning-oriented employees are intrinsicallymotivated to learn. A highly formalized team envi-ronment restricts the expression of this desire bylimiting discretion, thus attenuating the influenceof a learning orientation on creativity. In compari-son, learning-oriented employees working in lessrule bound team environments experience fewerrestrictions and have greater discretion to expresstheir inclination to try new approaches. Accord-ingly, as a learning orientation disposes individu-als to creatively engage with work problems andchallenges, we predict:

Hypothesis 2a. A learning orientation and for-malization interact to influence creativity; thelearning orientation is positively related to cre-ativity when formalization is low, but not whenformalization is high.

Performance-oriented individuals in teams withhigh formalization perceive clear guidelines as tothe behavior that is expected and seen as appropri-ate. Prove-oriented individuals use these guide-lines to regulate their behavior, which results inrule following and unwillingness to engage in be-haviors deviating from these standard procedures.In effect, this reduces prove-oriented individual’screative tendencies. Teams with little formalizationof rules and procedures, on the other hand, providean environment in which there is less enforcedclarity as to desired ways to engage with the job.Prove-oriented individuals seeking to demonstratehigh performance in a context with few formalrules and procedures thus have little information asto how they can demonstrate competence in rela-tion to their peers other than by getting the jobdone, and done well. Compared to situations ofhigh formalization, this situation may invite morecreativity in meeting work challenges from individ-uals with a stronger motivation to demonstrate

their competence. Developing creative solutions toproblems others have failed to resolve provides aparticularly powerful demonstration of compe-tence and so provides an opportunistic chance todemonstrate one’s capability, encouraging prove-oriented employees to engage in creative problem-solving behaviors.

Hypothesis 2b. A prove orientation and formal-ization interact to influence creativity; theprove orientation is positively related to cre-ativity when formalization is low, but not whenformalization is high.

For an avoid orientation, too, there is reason toexpect a less evident influence in teams withhigher formalization of practices. The reason forthis is the same as it is for learning and proveorientations: The more rules and procedures guidejob performance, the less freedom there is for theexpression of individual differences. In teams lowin formalization, however, the absence of pre-scribed structure and the associated freedom mayincrease uncertainty about the appropriate ways toengage with the job. Avoid-oriented individualstend to avoid challenges that carry the risk of errorsand failure and to favor endeavors with a highchance of success (VandeWalle, 1997). Whenguidelines are few, the risk of failure may loomlarge and invite avoid-oriented individuals to steerclear of creative challenges. Thus, whereas formal-ized procedures may do little to promote creativity,they at least work to keep avoid-oriented individ-uals’ tendency to avoid creative challenges incheck, rendering the relationship between an avoidorientation and creativity weaker under conditionsof high formalization than under conditions of lowformalization. Note that this is where the differencebetween low centralization and low formalizationdiscussed earlier expresses itself. Whereas decen-tralization in fact is an active influence that makesthe avoid orientation–creativity relationship weakerthan it is under conditions of high centralization (Hy-pothesis 1c), low formalization merely reflects theabsence of rules and procedures regulating behaviorand actually provides more opportunity for the ex-pression of an avoid orientation than high formaliza-tion. Thus, we hypothesize:

Hypothesis 2c. An avoid orientation and for-malization interact to influence creativity; theavoid orientation is negatively related to cre-ativity when formalization is low, but not whenformalization is high.

2011 629Hirst, van Knippenberg, Chen, and Sacramento

Page 7: HOW DOES BUREAUCRACY IMPACT INDIVIDUAL.pdf

METHODS

Research Setting, Participants, and Procedures

A national survey of employees of the TaiwanCustoms Bureau was conducted, comprising thoseworking at the Bureau’s offices at the internationalairports of Taipei and Kaohsiung and the interna-tional harbors of Keelung, Taichung, and Kaohsi-ung. Employees worked in teams that were respon-sible for a particular activity (e.g., developingfinance protocols or search procedures), workedrelatively independently of each other, and hadleaders with a high degree of influence on team-work outcomes. Forty-three percent of the teamsperformed operational search and detection tasks.Given the strategic sensitivity of the region and theneed to keep pace with a “technology and crimearms race” (Ball, 2003) in which the nature andmethods of offenses varied (from trafficking con-trolled substances, narcotics, animal and farm pro-duce, protected species, and organs and biomedicalby-products, to visa fraud and people smuggling),this was a particularly challenging function. Cre-ative problem solving was required to develop newinnovative detection approaches to keep pace withdiverse criminal activities. For example, creativitywas required to teach scent-detecting dogs trainedin another country to identify new scents associ-ated with offenses specific to the Taiwanese con-text (e.g., betel nut smuggling). The remaining 57percent of the teams performed administrative andmanagement functions such as performance monitor-ing, information technology, finance, and human re-sources activities. Given a strategic mandate to workand share data with other agencies, the Customs Bu-reau placed much emphasis on systems developmentand integration. For example, since the September 11,2001, attacks, Taiwan Customs have worked withU.S. government agencies to share security informa-tion, and this interaction has necessitated new com-puter systems as well as the development and appli-cation of new X-ray inspection technologies for cargoand containers. Thus, a combination of challengingdemands and limited resources made it essential foremployees to display creativity in developing inno-vative solutions to problems. Recognizing the neces-sity of creativity and innovation, management imple-mented a department-wide total quality management(TQM) initiative that included appraisals as well asmonetary rewards or credits to encourage employeesto submit creative ideas to improve work processes orsolve work problems. With these approaches, the Bu-reau sought to empower officers to streamline workpractices to reduce red tape, yet also assure theyworked within government regulations and demon-strated procedural compliance.

Employees’ creativity was appraised in annualperformance reviews and promotion decisions.Performance reviews included criteria assessinghow subordinates used creativity in their jobs (e.g.,by developing processes to improve work effi-ciency). Promotion decisions also assessed the“creativity of employees’ researching and work-ing.” For example, a subordinate grappling with anew data-collecting system would be appraised onwhether he or she had developed new ways toaddress or solve problems. Subordinates whoachieved positive assessments obtained credits intheir promotion evaluation.

The survey was translated by one of the authorsand two university faculty using the proceduresdescribed by Brislin (1980). At the project’s initia-tion, we conducted a small pilot test with 36 em-ployees to make sure all measures were reliable andused terms appropriate to the organization. Havingobtained adequate to good psychometrics and ver-bal feedback from participants, we slightly refinedthe translation of some of the measures for thelarger study. A rank-and-file, long-standing mem-ber of the organization who was supportive of thebroader project personally distributed a paper sur-vey to teams, ensuring a 100 percent completionrate. In total, 388 employees completed question-naires containing the independent variables, and 97team leaders rated employee creativity. As in previ-ous studies (Baer & Oldham, 2006; Zhou & Shalley,2008), supervisors were chosen to appraise creativity,as providing appraisals was part of their job. As such,we expected them to be in a prime position to eval-uate their employees and to be comfortable, willing,and fairly adept at doing so. Prior research supportsthe validity of this measurement approach. Supervi-sor ratings have been found to relate to creative out-put (e.g., suggestions for improvement schemes, in-vention disclosure forms and patents [Tierney,Farmer, & Graen, 1999]). Participants had the optionof responding anonymously, which precludedmatching questionnaires to supervisor ratings, and inaddition some records were obsolete, either becauseemployees had left the organization or administrativerecords were incomplete. We therefore obtainedmatched data for 330 employees and 95 teams, pro-viding a usable sample of 90 and 97 percent of theoriginal individual- and team-level populations.

The majority of sample members were male(73%); the average age was 44 years, and averagetenures in the organization and in their currentposition were 16 and 6 years. A quarter of thesample had been educated at junior college (theequivalent of a short, practically oriented univer-sity program); 64 percent had university qualifica-tions; and 10 percent had postgraduate degrees. Em-

630 JuneAcademy of Management Journal

Page 8: HOW DOES BUREAUCRACY IMPACT INDIVIDUAL.pdf

ployees were paid at varying salary “bands”according to their length of service. More senior em-ployees who were on higher bands tended to work inroles that provided greater decision-making latitude.

Measures

Goal orientation. We used VandeWalle’s (1997)three-factor scales to assess goal orientation as“learning,” “avoid,” or “prove.” The rating scale forall items ranged from 1 “not at all,” to 7 “to a largeextent.” An example of the five-item learning ori-entation scale is “I often look for opportunities todevelop new skills and knowledge.” An example ofthe four-item avoid orientation scale is “I’m con-cerned about taking on a task at work if my perfor-mance would reveal that I had low ability.” Anexample of the four-item prove orientation scale is“I’m concerned with showing that I can performbetter than my co-workers.”

Centralization. In selecting a centralization mea-sure, we also reviewed scales constructed for anorganizational context rather than a team context(e.g., Lee & Choi, 2003; Raub, 2007). Because teamcontext was a core element in both our theory de-velopment and research setting, however, we wereleft with the choice of adapting one of these scalesor selecting a team-focused alternative. Given thegood reliability and validity reported for it, weselected Arnold et al.’s (2000) four-item measure,which assesses the extent to which within-teamdecision making is centralized in the leader role asopposed to decentralized in participative decisionmaking. The items, rated from 1, “strongly disagree,”to 7, “strongly agree,” assess centralization of deci-sion making, using a team’s leader as the referent, asfollows: “Uses my work group’s suggestions to makedecisions that affect us,” “Listens to my work group’sideas and suggestions,” “Encourages work groupmembers to express ideas/suggestions,” and “Givesall work group members a chance to voice their opin-ions.” To facilitate interpretation, we reverse-scoredthe scale so that higher ratings reflected greater cen-tralization of within-team decision making.

Formalization. We used a three-item scale (Hage& Aiken, 1967; Rafferty & Griffin, 2004) relating tothe extent to which work was structured and regu-lated by rules and protocols to assess formalization.The items, rated on a scale ranging from 1,“strongly disagree,” to 5, “strongly agree,” were“There are a lot of rules and regulations in thisteam,” “Our work involves a great deal of paper-work and administration,” and “Our work is highlyregulated by bureaucratic procedures.”

Creativity. We measured employees’ creativityusing the four-item scale reported by Farmer, Tier-

ney, and Kung-McIntyre (2003). Team leaders ratedemployees’ creativity on a scale ranging from 1, “notat all correct,” to 6, “completely correct.” Items were“Seeks new ideas and ways to solve problems,” “Gen-erates ideas revolutionary to the field,” “Is a good rolemodel for innovation/creativity,” and “Tries newideas and approaches to problems.”

Control variables. We controlled for gender (0 �“male,” 1 � “female”), education (1, “high school,”to 5 � “postgraduate degree: master’s or Ph.D.”),individual age and mean team age, individual andteam tenure, and team size, as each has been foundto relate to employee creativity (Tierney & Farmer,2002) and to influence team processes (Hirst et al.,2009). We included additional controls to take intoaccount both the heterogeneity of the sample andpractices particular to the Customs Bureau. Wecontrolled for employees’ remuneration band (i.e.,more senior employees were generally paid moreand afforded greater discretion and so greater op-portunities for creativity). At the team level, wecontrolled for whether a team was “operational”(coded 1) or “administrative/managerial” (coded2). To control for potential differences in officesand practices, we created two dummy variables forregion: region 1 was Taipai, the capital and largestairport, and region 2 was Keelung, a distinctivelocale that hosts the most remote provincial office,which is responsible for large volumes of sea-basedtrade through Keelung Harbor.

Validation of Multilevel Data Structure

Analyses included individual- and team-levelconstructs. Performing two different analyses tovalidate this data structure, we first examinedwhether the data justified aggregation of team-levelconstructs. According to a one-way analysis of vari-ance (ANOVA), centralization and formalizationdiffered between teams (p � .05). Mean values ofrwg(j) across teams of .89 for centralization and .92for formalization, calculated using a normal distri-bution, suggested adequate within-team agreement(James, Demaree, & Wolf, 1984). Thus the rwg(j) sup-ported aggregation of constructs to the team level.For a more comprehensive test of the multileveldata structure, we conducted a multilevel factoranalysis following procedures recommended byDyer, Hanges, and Hall (2005). We first performedconfirmatory factor analysis (CFA) on the six con-structs of learning, prove, and avoid orientations,and centralization, formalization, and creativity.The fit of the model with six factors loading sepa-rately (�2 � 493.26, df � 237, p � .01, RMSEA �.06, CFI � .95) was compared with the fits of sixpotential alternatives. The first alternative was a

2011 631Hirst, van Knippenberg, Chen, and Sacramento

Page 9: HOW DOES BUREAUCRACY IMPACT INDIVIDUAL.pdf

five-factor model with prove and avoid orientationscollapsed to represent an overall performance ori-entation factor (�2 � 818.06, df � 242, p � .01,RMSEA � .09, CFI � .88), and the second alterna-tive model had the learning and prove orientationscollapsed into one “approach” factor (�2 �1,224.97, df � 242, p � .01, RMSEA � .11, CFI �.80). The third alternative was a four-factor modelcollapsing all three orientations into one goal ori-entation factor (�2 � 1,565.06, df � 246, p � .01,RMSEA � .13, CFI � .73); next, we tested a three-factor model also collapsing centralization and for-malization into one bureaucracy factor (�2 �1,664.81, df � 249, p � .01, RMSEA � .13, CFI �.71); a two-factor model collapsing all independentvariables into one factor (�2 � 2959.09, df � 251,p � .01, RMSEA � .18, CFI � .44); and a one-factormodel (�2 � 3,458.06, df � 252, p � .01, RMSEA �.20, CFI � .33). These findings show that the six-factor model provided a good fit with the data anda better fit than the alternative models.

These data modeled at the individual level thusdemonstrated the convergent and discriminant va-lidity of the constructs studied, providing suffi-cient basis to test the multilevel structure of thedata (Dyer et al., 2005). We expected that the six-factor structure of the model would be consistent atboth levels; thus we constructed within- and be-tween-team CFA models comprising six factors.Multilevel factor analysis is computationally de-manding, and convergence problems tend to arisewhen the number of observed indicators is high(Dyer et al., 2005). To avoid nonconvergence, weused item parcelling based on item skewness toreduce the number of observed indicators (Nasser &Wisenbaker, 2003). This procedure resulted in a re-duction in the number of indicators from 24 to 13 anda factor structure with acceptable fit (Muthen, 1994)at both the individual and group levels of analysis(�2 � 112.17, df � 100, p � .19, CFI � .99, RMSEA �.02, SRMRbetween � .06, SRMRwithin � .04).

RESULTS

Table 1 displays correlations among variables.Individual-level variables are below and aggregatedvariables above the diagonal. Our hypotheses arecross-level interaction hypotheses, and we usedhierarchical linear modeling (HLM) to test them. Fol-lowing Preacher, Curran, and Bauer’s (2006) recom-mendations, we group-mean-centered all individual-level (level 1) variables except for gender. Team-level(level 2) variables were not centered, to reduce pos-sible problems with multicollinearity (Raudenbush &Bryk, 2002). Table 2 summarizes the HLM results.First, we tested a null model into which no predictors

were entered. Next we introduced the individual-level variables (step 1), followed by the team variables(step 2), and in the final step we simultaneouslytested all cross-level interactions (step 3).

The null model allowed us to test the between-team variance in creativity by examining the level 2residual variance of the intercept (�00) and the ICC1statistic, which represents the proportion of vari-ance in the outcome variable that resided betweengroups. Analyses revealed a �00 of .14 (p � .001)and an ICC1 of .37, indicating 37 percent of thevariance resided between teams, to be explained bylevel 2 variables. A precondition for testing cross-level interactions here was that the slopes of rela-tions between the goal orientations and creativityvary across teams. This was the case (learning ori-entation: U1 variance � .14, �2[94] � 289.19, p �.01; prove orientation: U1 variance � .14, �2[94] �284.73, p � .001; avoid orientation; U1 variance �.14, �2[94] � 288.43, p � .01).

Main effects. The second and third sections ofTable 2 show the level 1 and 2 main effects. Of thelevel 1 control variables, only age was a significant,and negative, predictor of creativity (� � –.03, p �.05). Avoid orientation (� � –.08, p � .05) had anegative relation with creativity. None of the level2 variables were significantly related to employeecreativity. Note that although the graphs for theinteractions presented below might suggest maineffects for the dimensions of bureaucracy, thesewere not significant. These figures should be inter-preted in light of the fact that we followed Aikenand West’s (1991) method for examining interac-tions, whereby values are selected by convention(i.e., one standard deviation above and below themean). Were other values to be selected, the resultsmight differ, and so these figures should be inter-preted in terms of the slopes for the goal orienta-tions, and not the intercept for low and high bu-reaucracy. Moreover, we targeted these graphs toreflect one particular moderating influence on therelationship of one particular goal orientation withcreativity and did not fully factor in the influenceof the other model variables.

Cross-level interactions. We estimated slopes-as-outcomes models in HLM to assess the moderatingeffect of centralization and formalization on therelationship between goal orientation and em-ployee creativity. The final section of Table 2 pres-ents the results of this analysis. The model ex-plained 4 percent of the variance in creativity.Using the procedure described by Preacher et al.(2006) and software developed by Shacham (2009),we conducted simple slopes analysis for all cross-level interactions. Hypotheses 1a, 1b, and 1c pre-dict that centralization moderates the relations be-

632 JuneAcademy of Management Journal

Page 10: HOW DOES BUREAUCRACY IMPACT INDIVIDUAL.pdf

TA

BL

E1

Mea

ns,

Sta

nd

ard

Dev

iati

ons,

and

Cor

rela

tion

sa

Var

iabl

eM

ean

s.d

.1

23

45

67

89

1011

1213

1415

1617

1.G

end

er0.

310.

47�

.23

�.0

6�

.24*

*.1

0.0

0.0

3�

.07

�.4

2**

�.0

4.3

4**

�.2

3**

�.0

6�

.18

�.0

8.1

4�

.00

2.A

ge44

.01

8.05

�.2

1**

.87*

*.6

2**

�.3

6**

�.0

2.1

1�

.01

.15

�.2

7**

�.1

71.

00**

.87*

*.0

9�

.11

�.0

1�

.11

3.T

enu

re16

.14

9.17

�.0

6.8

9**

.62*

*�

.40*

*.0

0.1

1.0

7.0

3�

.32*

*�

.09

.87*

*1.

00**

�.0

4�

.12

.06

�.1

14.

Rev

enu

e82

.39

14.6

0�

.18*

*.6

3**

.66*

*�

.09

.09

.15

.11

.12

�.2

8**

�.1

2.6

2**

.62*

*.0

5�

.11

.06

�.0

45.

Ed

uca

tion

2.81

0.63

.03

�.3

2**

�.3

5**

�.0

6�

.05

.03

.10

�.0

9.1

4.1

8�

.36*

�.4

0**

�.0

1�

.05

�.0

4.0

56.

Lea

rnin

gor

ien

tati

on4.

791.

05.0

2.0

4.0

3.1

0.0

6(.

89)

�.2

0*.4

4**

.10

.14

.01

�.0

2�

.01

.06

�.1

4.1

2.0

47.

Avo

idor

ien

tati

on4.

121.

12�

.13*

.14*

*.0

9.1

4**

.04

.09

(.82

).2

2*�

.02

�.1

5.1

3.1

1.1

1.0

7.0

7�

.01

�.1

58.

Pro

veor

ien

tati

on4.

101.

22�

.14*

.17*

*.1

4**

.18*

*.0

2.4

7**

.47*

*(.

93)

.15

�.0

2�

.07

�.0

1.0

1�

.05

�.1

2�

.02

�.1

19.

Cat

egor

y1.

430.

50.2

6**

.12*

.05

.08

�.1

1.0

7.0

3.0

7.3

2**

�.2

8**

.15

.03

.21*

�.1

3.2

4*�

.15

10.

Reg

ion

10.

290.

46.0

1�

.19*

*�

.24*

*�

.17*

*.1

4*.0

7�

.00

.08

.32*

*�

.21

�.2

7**

�.3

2**

.17

�.0

5.0

9�

.03

11.

Reg

ion

20.

070.

26.1

9**

�.0

9�

.06

�.0

8.0

7.0

1�

.02

.07

�.2

8**

�.1

8**

�.1

7�

.09

�.0

4�

.15

�.0

1.1

212

.M

ean

age

43.9

94.

83�

.08

.34*

*.3

2**

.21*

*�

.20*

*�

.04

.03

.02

.09

�.2

6**

�.1

7**

.87*

*.0

9�

.11

�.0

1�

.11

13.

Tea

mte

nu

re15

.81

5.91

.00

.53*

*.6

3**

.39*

*�

.29*

*.0

1.0

3.0

5.0

7�

.38*

*�

.09

.52*

*�

.04

�.1

2�

.06

�.1

114

.T

eam

size

6.98

2.01

�.1

4*.0

3�

.04

.03

.01

.03

.06

�.0

4.2

0**

.08

�.0

4.0

8�

.06

.04

.13

�.1

715

.C

entr

aliz

atio

n2.

791.

07�

.03

�.0

2�

.06

�.0

6�

.02

�.1

9**

�.0

2�

.11*

�.1

0�

.06

�.0

8.1

0.0

4.0

4(.

99)

.03

�.0

516

.F

orm

aliz

atio

n3.

700.

62.0

6�

.10

�.0

4.0

1�

.01

.15*

*.0

4.1

1*.2

0**

.06

�.0

1�

.07

�.0

5.0

4.0

1(.

65)

.03

17.

Cre

ativ

ity

3.61

0.60

.01

.12*

�.0

8.0

2.0

9.1

1*�

.13*

�.0

4�

.10

.02

.08

�.0

8.1

2*.1

3*�

.04

.02

(.84

)

aIn

div

idu

al-l

evel

anal

yses

are

belo

wth

ed

iago

nal

,an

dgr

oup

-lev

elan

alys

esar

eab

ove

the

dia

gon

al.

Inte

rnal

con

sist

ency

reli

abil

itie

sar

ein

par

enth

eses

.n

�95

team

sco

mp

risi

ng

330

emp

loye

es.

*p

�.0

5**

p�

.01

Page 11: HOW DOES BUREAUCRACY IMPACT INDIVIDUAL.pdf

tween creativity and the learning, prove, and avoidorientations. The interaction of learning orientationand centralization was significant (� � –.06, p �.05). Figure 1A depicts this interaction. The slopewas significant and positive when centralizationwas low (one standard deviation below the mean;� � .12, t � 3.20, p � .01) and nonsignificant whenit was high (one standard deviation above themean; � � .01, t � .19, n.s.). These results supportHypothesis 1a.1 The prove orientation by central-

ization interaction was not significant, failing tosupport Hypothesis 1b. The interaction of theavoid orientation and centralization was significant(� � –.08, p � .05; see Figure 1B). The slope wassignificant and negative when centralization washigh (� � –.16, t � 3.74, p � .01) and nonsignifi-cant when centralization was low (� � –.01, t �.27, n.s.). These results support Hypothesis 1c.

Hypotheses 2a, 2b, and 2c predict that formal-ization moderates the relation between creativityand learning, prove, and avoid orientations. Not

1 Two recent studies have shown nonlinear relationsfor a learning goal orientation (Bunderson & Sutcliffe,2003; Hirst et al., 2009) and predictors of employee cre-ativity (Baer & Oldham, 2006). Therefore, although theserelations were not hypothesized here, we entered qua-dratic and cubic terms into cross-level interactions for

both team-level variables. For exploratory purposes andin anticipation of readers’ potential interest, we alsotested three-way interactions reflecting possible interac-tive effects of centralization and formalization. None ofthese interactions were significant.

TABLE 2HLM Results for the Effects of Cross-Level Interactions of Goal Orientation

with Team Bureaucracy on Employee Creativity

Variable Coefficient s.e. t �2 Model Deviance R2b Rtotal2c

Null modelIntercept 3.62* .05 77.09* 285.33 585.15

Level 1 variablesIntercept 3.63* .05 78.13* 304.64 584.49 .02Gender �0.09 .08 �1.24Age �0.03* .01 �2.47*Tenure 0.01 .01 1.37Revenue 0.01 .00 1.70Education 0.03 .06 0.50Learning 0.05 .03 1.49Prove 0.03 .03 0.95Avoid �0.08* .03 �2.28*

Level 2 main effectsCategory �0.07 .12 �0.63 281.20 612.31 .00 .02Region 1 (two offices) 0.01 .12 0.10Region 2 (four offices) 0.14 .18 0.82Mean age 0.01 .01 0.61Team tenure �0.02 .01 �1.74Team size �0.04 .02 �1.66Centralization �0.02 .04 �0.35Formalization 0.01 .05 0.19

All interaction termsLearning orientation � centralization �0.06* .03 �1.97* 297.92 623.12 .03 .04Prove orientation � centralization 0.03 .03 1.11Avoid orientation � centralization �0.08* .03 �2.58*Learning orientation � formalization �0.02 .04 �0.59Prove orientation � formalization �0.10* .04 �2.70*Avoid orientation � formalization 0.08* .03 2.26*

a Employee n � 330, team n � 95.b Indicates the proportion of variance explained at each level; i.e., level 1 within-team variance, level 2 between-team variance and

cross-level interactions.c R2

total � R2within-group � (I � ICC1) � R2

between-groups � ICC1.* p � .05

Two-tailed tests.

634 JuneAcademy of Management Journal

Page 12: HOW DOES BUREAUCRACY IMPACT INDIVIDUAL.pdf

supporting Hypothesis 2a, the learning orienta-tion by formalization interaction was not signif-icant. The interaction of a prove orientation andformalization was significant (� � –.10, p � .05;see Figure 2A). The slope was positive and sig-nificant when formalization was low (� � .11, t �2.87, p � .01) and negative and nonsignificantwhen it was high (� � –.06, t � 1.03, n.s.). Theseresults support Hypothesis 2b. The interaction ofthe avoid orientation and formalization was sig-nificant (� � .08, p � .05; see Figure 2B). The

slope was significant and negative when formal-ization was low (� � –.14, t � 3.42, p � .01) andnonsignificant when it was high (� � –.00, t �0.03, n.s.). These results support Hypothesis 2c.

DISCUSSION

Managers face the challenge of getting creativeresults from the individuals in their teams, yet adesire for control and behavioral regulation mayinspire centralized decision making and formal-

FIGURE 1Interactions for Centralization

3.5

4

4.5

5

(1A) Interaction of Centralization and Learning Orientation Predicting Creativity

(1B) Interaction of Centralization and Avoid Orientation Predicting Creativity

Low High

EmployeeCreativity

High centralization

Low centralization

Learning Orientation

3.5

4

4.5

5

Low High

EmployeeCreativity

Avoid Orientation

2011 635Hirst, van Knippenberg, Chen, and Sacramento

Page 13: HOW DOES BUREAUCRACY IMPACT INDIVIDUAL.pdf

ization of rules and procedures. Our findingsconfirm that team bureaucracy can suppress theexpression of individual differences that may en-gender creativity. A learning orientation was pos-itively related to creativity only under conditionsof low centralization, and a prove orientation waspositively related to creativity only when formal-ization was low. Moreover, low centralizationattenuated the negative association between anavoid orientation and creativity. Although theevidence would seem to favor lower bureaucracyfor engendering creativity, findings for the avoidorientation by formalization interaction providean important caveat: less formalized practiceswere associated with a more negative relation-

ship between an avoid orientation and creativity.These findings illustrate that different goal orien-tations produce different creative responses to agiven context and suggest the promise of person-in-situation perspectives for understanding teambureaucratic influences.

Theoretical Implications

Although it is well established that both individ-ual differences and context play an important rolein the creative process (Amabile, 1996; Shalley etal., 2004), person-in-situation approaches that ac-count for their interactive influences are still indevelopment. Goal orientations are of particular

FIGURE 2Interactions for Formalization

3.5

4

4.5

5

Low High

High formalization

Low formalization

Prove Orientation

EmployeeCreativity

(2A) Interaction of Formalization and Prove Orientation Predicting Creativity

(2B) Interaction of Formalization and Avoid Orientation Predicting Creativity

3.5

4

4.5

5

Low High

Avoid Orientation

Employee Creativity

636 JuneAcademy of Management Journal

Page 14: HOW DOES BUREAUCRACY IMPACT INDIVIDUAL.pdf

interest in this respect, given their strong links withbehavior in achievement situations and reviewshighlighting the need to examine goal orientationsas they unfold dynamically in context (Button etal., 1996; DeShon & Gillespie, 2005). In an impor-tant counterpoint to earlier work by Hirst et al.(2009) demonstrating that team contextual influ-ences may stimulate the creative expression of goalorientations, the present study shows that the teamcontext may inhibit as well as stimulate the linkbetween goal orientations and creativity—and,moreover, that this inhibition is not necessarily badfor creativity. These findings are important in sug-gesting that the influence of a situation on the cre-ative expression of dispositions should be under-stood not only in terms of the extent to which thesituation provides cues that may activate the ex-pression of traits, but also in terms of the extent towhich it constrains individuals’ behavior and thusinhibits the expression of individual differences.This argument suggests the contours of a more com-prehensive model of team contextual influences onthe relationship between individual differencesand creativity, a model that incorporates both trait-activating and trait-inhibiting influences. Althoughsuch a model is clearly consistent with trait activa-tion theory (Tett & Burnett, 2003), the obvious chal-lenge for creativity research is to identify specifictrait-context combinations that are relevant to cre-ativity. An important issue to take into account inthis respect is that the present findings for central-ization show that the exact same contextual cuemay activate one trait while inhibiting another.

For research in creativity, the findings for teambureaucracy are important in that they show thatthe creativity of those motivated to engage withwork challenges (i.e., learning- and prove-orientedindividuals) benefits from low bureaucracy. It isless clear why, for learning-oriented people, thisrelationship holds for low centralization but notlow formalization, whereas for prove-oriented in-dividuals, it would hold for low formalization butnot for low centralization. We are hesitant to baseconclusions on null findings (i.e., the absence oflearning orientation by formalization and prove ori-entation by centralization interactions). Consis-tently with predictions, the pattern of results ismore complex for the avoid orientation and bureau-cracy. Decentralized decision making helped atten-uate the negative relationship between the avoidorientation and creativity, but low formalizationactually brought out this relationship. As we out-lined in the introduction, this can be seen as re-flecting an important difference between the activeand engaging nature of low centralization as com-pared with the passive nature of low formalization.

Our findings may also provide important pointersfor identifying further ways to manage the undesir-able outcomes of an avoid orientation (Dweck,1999). Encouraging avoid-oriented individuals toparticipate makes it more likely that they will over-come their disposition to avoid job challenges.

Findings for team bureaucracy also speak to thevalue of taking lower levels of analyses into ac-count than is typically done in research on bureau-cracy and related governance issues (Bolin &Harenstam, 2008; Caruana et al., 1998; Hall, 1999).To reconcile the tension between bureaucracy andcreativity, individual differences should be consid-ered as important moderators of the effectiveness oforganizational practices. In other words, these indi-vidual dispositions provide insight as to why differ-ent people respond so differently to the same bureau-cratic experiences. This in turn sheds light on how todevelop more constructive solutions to the ten-sion between managerial control and innovation.

We have pointed out that the available evidencefor goal orientation–creativity relationships sug-gests that these should be expected to manifest notas main effects, but rather, as contingent on contex-tual influences. The present findings by and largecorroborate this analysis. Only the avoid orienta-tion had a main effect; the learning and prove ori-entations did not. As the evidence for the contin-gent nature of the goal orientation–creativityrelationship amasses, it is interesting to note thatmain effects of goal orientations for in-role taskperformance have been more consistently docu-mented (Porath & Bateman, 1996; VandeWalle,Brown, Cron, & Slocum, 1999). This could reflectfundamental differences between creativity and in-role performance that render creativity—or at leastthe expression of individual differences in creativi-ty—inherently more contingent on contextual influ-ences. Future research that combines the study ofcreativity and in-role performance may be highlyworthwhile as a way to shed more light on this issue.

The systematic development of the goal orienta-tion literature has been hampered by differing con-ceptualizations of the construct and the wide arrayof scales in use; no common definition or set ofmeasures of goal orientation exists (DeShon & Gil-lespie, 2005). Our framework (cf. Elliot, 1999; Elliot& Church, 1997) relies on a three-dimensional con-ceptualization (Seijts et al., 2004; VandeWalle,1997) that partitions performance orientation intoits prove and avoid components. Although this isthe most widely accepted conceptualization of goalorientation, we recognize that an earlier two-factorconceptualization grouped the prove and avoid ori-entations under a general performance orientationconstruct (Button et al., 1996; Stevens & Gist, 1997).

2011 637Hirst, van Knippenberg, Chen, and Sacramento

Page 15: HOW DOES BUREAUCRACY IMPACT INDIVIDUAL.pdf

In this respect, our findings clearly testify to thevalidity and “value-added” of the distinction be-tween prove and avoid performance orientations.CFA showed that these represent two distinct fac-tors, and relationships with creativity clearly di-verge for prove and avoid orientations. We notethat on similar grounds, our findings also speakagainst grouping learning and prove orientationsunder a more general “approach motivation” con-struct (cf. Elliot, 1999). CFA demonstrated thatthese are distinct factors, and relationships withcreativity also diverged. The current three-dimen-sional goal orientation framework thus can be seenas providing the most valid basis for the interpre-tation of our results.

Managerial Implications

Bureaucratic practices often hinder managers’ ef-forts to facilitate individual initiative and creativity(Bolin & Harenstam, 2008). To get creative results,managers should seek to understand employees’motivational orientation in context, so the combi-nation of individual disposition and bureaucraticcontext yields the most desirable associations withcreativity. Decentralization brings out the best inlearning-oriented employees and attenuates thenegative effects of an avoid orientation on creativ-ity. Thus, from a creativity perspective, decentral-ized decision making is attractive, even when itdoes little for prove-oriented individuals.

Formalized practices are a more complex issue.Although the creativity of prove-oriented individ-uals thrives in less formalized contexts, low formal-ization actually brings out undesirable tendenciesin avoid-oriented team members. Given its verynature, formalization cannot be tailored to individ-ual team members, and these findings rather speakto the issue of selection and person-environmentfit. An avoid orientation brings few creative bene-fits. At best, it is not negatively related to creativity,but achieving such an absence of association relieson formalizing procedures at the expense of thecreative benefits of a prove orientation. Accord-ingly, the conclusion would be that creativity isbest obtained by (where possible) not selectingavoid-oriented individuals for a team, while creat-ing a team context characterized by high levels ofdecentralized decision making and low formaliza-tion, to let learning-oriented and prove-orientedindividuals flourish.

Limitations and Directions for Future Research

We tested how individuals responded to a bu-reaucratic context but did not test the mediating

processes underpinning these relationships. In thisrespect, the goal orientation framework suggeststhat different goal orientations lead individuals torespond to different aspects of the same situation.Whereas learning-oriented individuals may re-spond to challenges and opportunities for learningand further development, prove-oriented individu-als may respond primarily to cues that identifyopportunities to display their competence. Avoid-oriented individuals, in contrast, seem more drivenby cues highlighting the safety or potential risksassociated with engaging in certain actions in termsof how others will assess their performance. Futureresearch tapping perceptions of situations alongthose lines as potentially mediating processes mayfurther validate the current cross-level perspectiveon individual creativity. It may also provide impor-tant information for interpreting our less expectedfindings, such as the absences of a learning orien-tation by formalization interaction and of a proveorientation by centralization one. It is possible thatsituations of low centralization and low formaliza-tion, although both reflecting low bureaucracy, dif-fer in the eyes of learning-oriented and prove-ori-ented individuals as to the key triggers of theirmotivational drives (e.g., for learning-oriented in-dividuals, decentralization may be more associatedthan low formalization with learning opportunities).

We examined antecedents of employee creativityand did not test whether these effects generalized toin-role performance (i.e., activities mandated in anemployee’s position description and work designa-tion). We highlight this is as a much-needed area forfuture work, as creativity and performance research-ers have studied similar constructs but conductedfew comparisons between the two. Research examin-ing both outcomes would help explain whether dif-ferent antecedents are required to stimulate them andthus would allow tests of the generalizability of dif-ferent theoretical frameworks. We also suggest thereis a need to understand how cultural values impingeon the utility of goal orientation theory, which isderived from European-American values. Results forEast Asian samples, such as the correlations betweenlearning orientation and creativity observed in thecurrent study (r � .11) and by Gong et al. (2009; r �.20), are consistent with Harrison, Neff, Schwall, andZhao’s (2006) results in a meta-analysis of European-American research (range of r � .09–.30; mean r �.17). Yet one can speculate as to how cultural valuesmay also influence the predicted relationships. Ex-tending the current framework, we suggest that a cul-ture such as Taiwan, characterized by higher uncer-tainty avoidance than the United States, may be morelikely to invite the expression of an avoid orientationat the expense of creativity. That is, all other things

638 JuneAcademy of Management Journal

Page 16: HOW DOES BUREAUCRACY IMPACT INDIVIDUAL.pdf

being equal, the avoid orientation may have a stron-ger impact on creativity in, for instance, Taiwan thanin the U.S. We also see parallels between centraliza-tion and “power distance” (Hofstede, Neuijen,Ohayv, & Sanders, 1990) that would suggest that, allelse being equal, low power distance cultures aremore likely to invite creativity based on a learningorientation. Similarly, the more collectivistic orienta-tion of a country such as Taiwan as compared to, forinstance, the U.S., may also play a role. A perfor-mance approach orientation has a strong connotationof individual competition, which may be more easilyactivated in an individualistic than a collectivisticculture. The more general point is that culture can beseen as context too, and as such cultural context mayinfluence goal orientation–creativity relationships.Expanding knowledge of different indigenous con-texts and thus providing globally informed perspec-tives is important for management research if schol-ars are to speak to international business andcommerce trends—particularly as East Asian econo-mies look set to overtake established European andAmerican economies in the next decades.

We note that effect sizes for the interactions weresmall. An important issue in this respect is that sur-vey research can be assumed to lead to underestimat-ing the effect size of interactions (Evans, 1985;McClelland & Judd, 1993). This not only means that itis difficult to detect interactions in survey research(cf. Harrison, Price, & Bell, 1998), but also that effectsize may not be a fair criterion for judging them(Evans, 1985). Evans suggested that, given this under-estimation problem, interactions with explainedvariance of as little as 1 percent should be takenseriously, as they most likely represent larger trueeffect sizes. In view of these considerations as wellas more general evidence that goal orientations maynot have simple direct relationships with creativity(e.g., Hirst et al., 2009), we would argue that estab-lishing the current moderated relationships is im-portant to understanding creativity at work. Evenwhen explained variance appears to be small, itbuilds fundamental theory to understand the teamcontingencies relevant to the expression of individ-ual differences in creativity, and such understand-ing represents an important step toward identifyingthe conditions under which goal orientations aremost relevant to creativity.

Conclusion

Our study shows that team bureaucratic practicesmay regulate and influence individuals’ goal-directedbehavior to activate as well as inhibit the creativeexpression of goal orientations. In this respect, decen-tralization delivers the greater creative benefits, invit-

ing creativity from those intrinsically inclined tolearn and attenuating individuals’ avoidant tenden-cies. This is not to say that bureaucratic practicesalways diminish creativity; indeed, they may alsocurb less desirable avoidant tendencies. As bureau-cratic practices do little to bring out creativity, thechallenge is therefore to encourage creativity by re-ducing centralization and formalization, while notinviting individuals’ creativity-avoidant tendencies.

REFERENCES

Adler, P. S. 1999. Building better bureaucracies. Acad-emy of Management Executive, 13(4): 36–47.

Ahearne, M., Mathieu, J., & Rapp, A. 2005. To empoweror not to empower your sales force? An empiricalexamination of the influence of leadership empower-ment behavior on customer satisfaction and perfor-mance. Journal of Applied Psychology, 90: 945–955.

Aiken, L. S., & West, S. G. 1991. Multiple regression:Testing and interpreting interactions. NewburyPark, CA: Sage.

Amabile, T. M. 1988. A model of creativity and innova-tion in organizations. In B. M. Staw & L. L. Cum-mings (Eds.), Research in organizational behavior,vol. 10: 123–167. Greenwich, CT: JAI.

Amabile, T. M. 1996. Creativity in context: Update tothe social psychology of creativity. Boulder, CO:Westview.

Amabile, T. M., & Conti, R. 1999. Changes in the workenvironment for creativity during downsizing.Academy of Management Journal, 42: 630–640.

Arnold, J. A., Arad, S., Rhoades, J. A., & Drasgow, F.2000. The empowering leadership questionnaire:The construction and validation of a new scale formeasuring leader behaviors. Journal of Organiza-tional Behavior, 21: 249–269.

Baer, M., & Oldham, G. R. 2006. The curvilinear relationbetween experienced creative time pressure and cre-ativity: Moderating effects of openness to experienceand support for creativity. Journal of Applied Psy-chology, 91: 963–970.

Ball, D. 2003. Security trends in the Asia-Pacific re-gion: An emerging complex arms race. Workingpaper no. 380, Strategic and Defence Studies Centre,Australian National University.

Bolin, M., & Harenstam, A. 2008. An empirical study ofbureaucratic and post-bureaucratic characteristics in90 workplaces. Economic and Industrial Democ-racy, 29: 541–564.

Brislin, R. W. 1980. Translation and content analysis oforal and written material. In H. C. Triandis & J. W.Berry (Eds.), Handbook of cross-cultural psychol-ogy, vol. 1: 389–444. Boston: Allyn & Bacon.

Bunderson, J. S., & Boumgarden, P. 2010. Structure andlearning in self-managed teams: Why “bureaucratic”

2011 639Hirst, van Knippenberg, Chen, and Sacramento

Page 17: HOW DOES BUREAUCRACY IMPACT INDIVIDUAL.pdf

teams can be better learners. Organization Science,21: 609–624.

Bunderson, J. S., & Sutcliffe, K. M. 2003. Managementteam learning orientation and business unit perfor-mance. Journal of Applied Psychology, 88: 552–560.

Burns, T., & Stalker, G. M. 1961. The management ofinnovation. London: Tavistock.

Button, S. B., Mathieu, J. E., & Zajac, D. M. 1996. Goalorientation in organizational research: A conceptualand empirical foundation. Organizational Behaviorand Human Decision Processes, 67: 26–48.

Caruana, A., Morris, M., & Vella, A. 1998. The effect ofcentralization and formalization on entrepreneur-ship in export firms. Journal of Small BusinessManagement, 36: 16–29.

Chen, G., & Kanfer, R. 2006. Toward a systems theory ofmotivated behavior in work teams. In B. Staw &R. M. Kramer (Eds.), Research in organizationalbehavior, vol. 27: 223–267. Greenwich, CT: JAI.

Chen, G., Kirkman, B. L., Kanfer, R., Allen, D., & Rosen,B. 2007. A multilevel study of leadership, empower-ment, and performance in teams. Journal of AppliedPsychology, 92: 331–346.

DeShon, R. P., & Gillespie, J. Z. 2005. A motivated actiontheory account of goal orientation. Journal of Ap-plied Psychology, 90: 1096–1127.

Dweck, C. S. 1999. Self-theories: Their role in motiva-tion, personality and development. Ann Arbor, MI:Psychology Press.

Dyer, N. G., Hanges, P. J., & Hall, R. J. 2005. Applyingmultilevel confirmatory factor analysis techniques tothe study of leadership. Leadership Quarterly, 16:149–167.

Elliot, A. J. 1999. Approach and avoidance motivationand achievement goals. Educational Psychologist,34: 169–189.

Elliot, A. J., & Church, M. A. 1997. A hierarchical modelof approach and avoidance achievement motivation.Journal of Personality and Social Psychology, 72:218–232.

Evans, M. G. 1985. A Monte-Carlo study of the effects ofcorrelated method variance in moderated multiple-regression analysis. Organizational Behavior andHuman Decision Processes, 36: 305–323.

Farmer, S. M., Tierney, P., & Kung-McIntyre, K. 2003.Employee creativity in Taiwan: An application ofrole identity theory. Academy of Management Jour-nal, 46: 618–630.

Gong, Y., Huang, J. C., & Farh, J. L. 2009. Employeelearning orientation, transformational leadership,and employee creativity: The mediating role of em-ployee creative self-efficacy. Academy of Manage-ment Journal, 52: 765–778.

Hage, J., & Aiken, M. 1967. Relationship of centralization

to other structural properties. Administrative Sci-ence Quarterly, 12: 72–92.

Hall, R. D. 1999. Organizations: Structures, processes,and outcomes. London: Prentice-Hall.

Harrison, D. A., Price, K. H., & Bell, M. P. 1998. Beyondrelational demography: Time and the effects of sur-face- and deep-level diversity on work group cohe-sion. Academy of Management Journal, 41: 96–107.

Harrison, M. M., Neff, N. L., Schwall, A. R., & Zhao, X.2006. A meta-analytic investigation of individualcreativity and innovation. Paper presented at the 21stAnnual Conference of the Society for Industrial andOrganizational Psychology, Dallas.

Hirst, G., van Knippenberg, D., & Zhou, J. 2009. A cross-level perspective on employee creativity: Goal orienta-tion, team learning behavior, and individual creativity.Academy of Management Journal, 52: 280–293.

Hofstede, G., Neuijen, B., Ohayv, D. D., & Sanders, G.1990. Measuring organizational cultures—A qualita-tive and quantitative study across 20 cases. Admin-istrative Science Quarterly, 35: 286–316.

James, L. R., Demaree, R. B., & Wolf, G. 1984. Estimatingwithin group interrater reliability with and withoutresponse bias. Journal of Applied Psychology, 69:8–98.

Janssen, O., & Van Yperen, N. W. 2004. Employees’ goalorientations, the quality of leader-member exchange,and the outcomes of job performance and job satis-faction. Academy of Management Journal, 47:368–384.

Kirkman, B. L., & Rosen, B. 1997. A model of work teamempowerment. In R. W. Woodman & W. A. Pasmore(Eds.), Research in organizational change and de-velopment, vol. 10: 131–167. Greenwich, CT: JAI.

Klein, K. J., & Kozlowski, S. W. J. 2000. Multilevel the-ory, research, and methods in organizations. SanFrancisco: Jossey-Bass.

Kristof, A. L. 1996. Person-organization fit: An integra-tive review of its conceptualization, measurement,and implications. Personnel Psychology, 49: 1–49.

Lee, H., & Choi, B. 2003. Knowledge management en-ablers, processes, and organizational performance:An integrative view and empirical examination.Journal of Management Information Systems,20(1): 179–228.

McClelland, G. H., & Judd, C. M. 1993. Statistical diffi-culties of detecting interactions and moderator ef-fects. Psychological Bulletin, 114: 376–390.

Mischel, W. 1977. The interaction of person and situa-tion. In D. Magnusson & N. S. Endler (Eds.), Person-ality at the crossroads: Current issues in interac-tional psychology: 333–352. Hillsdale, NJ: Erlbaum.

Muthen, B. 1994. Multilevel covariance structure analy-sis. Sociological Methods & Research, 22: 376–398.

640 JuneAcademy of Management Journal

Page 18: HOW DOES BUREAUCRACY IMPACT INDIVIDUAL.pdf

Nasser, F., & Wisenbaker, J. A. 2003. A Monte Carlo studyinvestigating the impact of item parcelling on mea-sures of fit in confirmatory factor analysis. Educa-tional and Psychological Measurement, 63: 729–757.

Oldham, G. R., & Cummings, A. 1996. Employee creativ-ity: Personal and contextual factors at work. Acad-emy of Management Journal, 39: 607–634.

Porath, C. L., & Bateman, T. S. 2006. Self-regulation:From goal orientation to job performance. Journal ofApplied Psychology, 91: 185–192.

Preacher, K. J., Curran, P. J., & Bauer, D. J. 2006. Compu-tational tools for probing interactions in multiplelinear regression, multilevel modeling, and latentcurve analysis. Journal of Educational and Behav-ioral Statistics, 31: 437–448.

Rafferty, A. E., & Griffin, M. A. 2004. Dimensions oftransformational leadership: Conceptual and empir-ical extensions. Leadership Quarterly, 15: 329–354.

Raub, S. 2007. Does bureaucracy kill individual initia-tive? The impact of structure on organizational citi-zenship behavior in the hospitality industry. Inter-national Journal of Hospitality Management, 27:179–186.

Raudenbush, S. W., & Bryk, A. S. 2002. Hierarchicallinear models. Thousand Oaks, CA: Sage.

Seijts, G. H., Latham, G. P., Tasa, K., & Latham, B. W.2004. Goal setting and goal orientation: An integra-tion of two different yet related literatures. Academyof Management Journal, 47: 227–239.

Shacham, R. 2009. A utility for exploring HLM 2 and 3way interactions Software available from http://www.people.ku.edu/�preacher/interact/shacham/index.htm.

Shalley, C. E., Zhou, J., & Oldham, G. R. 2004. The effectsof personal and contextual characteristics on creativ-ity: Where should we go from here? Journal of Man-agement, 30: 933–958.

Srivastava, A., Bartol, K. M., & Locke, E. A. 2006. Em-powering leadership in management teams: Effectson knowledge sharing, efficacy, and performance.Academy of Management Journal, 49: 1239–1251.

Stevens, C. K., & Gist, M. E. 1997. Effects of self-efficacyand goal-orientation training on effective negotiationskills maintenance: What are the mechanisms? Per-sonnel Psychology, 50: 955–978.

Taggar, S. 2002. Individual creativity and group ability toutilize individual creative resources: A multilevelmodel. Academy of Management Journal, 45: 315–330.

Tett, R. P., & Burnett, D. D. 2003. A personality trait-based interactionist model of job performance. Jour-nal of Applied Psychology, 88: 500–517.

Thompson, V. A. 1965. Bureaucracy and innovation. Ad-ministrative Science Quarterly, 10: 1–20.

Tierney, P., & Farmer, S. M. 2002. Creative self-efficacy:Its potential antecedents and relationship to creativeperformance. Academy of Management Journal,45: 1137.

Tierney, P., Farmer, S. M., & Graen, G. B. 1999. Anexamination of leadership and employee creativity:The relevance of traits and relationships. PersonnelPsychology, 52: 591–620.

Van de Ven, A. H., & Ferry, D. L. 1980. Measuring andassessing organizations. New York: Wiley.

VandeWalle, D. 1997. Development and validation of awork domain goal orientation instrument. Educa-tional and Psychological Measurement, 57: 995–1015.

VandeWalle, D., Brown, S. P., Cron, W. L., & Slocum,J. W. 1999. The influence of goal orientation andself-regulation tactics on sales performance: A lon-gitudinal field test. Journal of Applied Psychology,84: 249–259.

Wagner, J. A. 1994. Participation’s effects on perfor-mance and satisfaction: A reconsideration of the re-search evidence. Academy of Management Review,19: 312–330.

Zhou, J., & Shalley, C. E. 2008. Handbook of organiza-tional creativity. New York: Erlbaum.

Giles Hirst ([email protected]) is an associate pro-fessor and the deputy director of research at MonashUniversity, Department of Management, Faculty of Busi-ness & Economics. He received his Ph.D. from the Mel-bourne Business School. His research interests includethe study of employee creativity and innovation.

Daan Van Knippenberg ([email protected]) is aprofessor of organizational behavior, Rotterdam Schoolof Management, Erasmus University Rotterdam. He re-ceived his Ph.D. from Leiden University. His researchinterests include leadership, diversity, team perfor-mance, creativity, and social identity.

Chin-Hui Chen ([email protected]) is acustoms officer at the Kaohsiung Customs Bureau, Tai-wan. He is also a part-time lecturer in management at theFooyin University. He received his Ph.D. in human re-source management from the National Sun Yat-Sen Uni-versity. His research interests include job stress, self-efficacy, psychological well-being, and goal orientation.

Claudia A. Sacramento ([email protected]) isan assistant professor of organizational behavior at theDepartment of Work and Organisational Psychology, As-ton Business School. She received her Ph.D. from AstonUniversity. Her research interests include creativity andteam effectiveness.

2011 641Hirst, van Knippenberg, Chen, and Sacramento

Page 19: HOW DOES BUREAUCRACY IMPACT INDIVIDUAL.pdf

Copyright of Academy of Management Journal is the property of Academy of Management and its content may

not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written

permission. However, users may print, download, or email articles for individual use.


Recommended