Academy of Management Journal 2011, Vol. 54, No. 3, 624641.

HOW DOES BUREAUCRACY IMPACT INDIVIDUAL CREATIVITY? A CROSS-LEVEL INVESTIGATION OF TEAM CONTEXTUAL INFLUENCES ON GOAL ORIENTATIONCREATIVITY RELATIONSHIPS
GILES HIRST Monash University DAAN VAN KNIPPENBERG Erasmus University Rotterdam CHIN-HUI CHEN Taiwan Customs Bureau CLAUDIA A. SACRAMENTO Aston University
Offering important counterpoint to work identifying team influences stimulating creative expression of individual differences in goal orientation, we develop cross-level theory establishing that team bureaucratic practices (centralization and formalization) constrain creative expression. Speaking to the tension between bureaucracy and creativity, findings indicate that this influence is not only negative and that effects of centralization and formalization differ. Surveying 330 employees in 95 teams at the Taiwan Customs Bureau, we found that learning and performance avoid goal orientations had, respectively, stronger positive and weaker negative relationships with creativity under low centralization. A performance-prove orientation was positively related to creativity under low formalization.

As employee creativity is crucial for organizational innovation and survival (Amabile, 1988; Oldham & Cummings, 1996), managers and scholars alike have sought to identify the ingredients that foster individual creativity. It is well recognized that the team context in which employees are embedded plays a central role in stimulating the creative expression of individual differences (Amabile & Conti, 1999; Hirst, van Knippenberg, & Zhou, 2009; Shalley, Zhou, & Oldham, 2004). Accordingly, researchers have begun to adopt a cross-level focus to examine the interplay between individual and team factors (Zhou & Shalley, 2008; cf. Klein & Kozlowski, 2000). In this respect, research has identified individual differences in goal orientation

The authors would like to thank Associate Editor Elizabeth Morrison and the three anonymous reviewers for their insightful comments and suggestions. We also thank Jeremy Dawson, Pamela Tierney, and Adam Grant for their advice and The Faculty of Business and Economics, University of Melbourne, for the support and facilities provided during the first authors sabbatical. Editors Note: The manuscript for this article was accepted during Duane Irelands term as editor.
624

that encourage self-regulation in achievement situations as a powerful influence on creativity when a context stimulates their expression (Hirst et al., 2009). Yet researchers have in effect turned a blind eye to the fact that organizations, and units within them, also need to impose practices and procedures that themselves regulate, order, and control behavior (Burns & Stalker, 1961; Thompson, 1965). Organizations and organizational units instill such practices to ensure consistency, efficiency, and control (Adler, 1999), but such bureaucracy may stifle individuals creativity. To build toward a comprehensive understanding of the factors affecting individual creativity in teams, scholars thus need to consider not only contextual factors that invite creativity, but also contextual influences that may constrain it. To provide this important complementary perspective on the current state of the science, we zoom in on team bureaucratic context and how it influences the creative expression of goal orientations. We rely on an integrative person-in-situation theory that describes how situational influences may either restrain or invite the expression of individual differences (Tett & Burnett, 2003; cf. Mischel, 1977). Although at first glance it might appear that

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 holders express written permission. Users may print, download or email articles for individual use only.

2011

Hirst, van Knippenberg, Chen, and Sacramento

625

the influence of bureaucracy is straightforwardit suppresses creativitywe propose that a closer look suggests that the issue is more complex. First, there are important individual differences in goal orientations that capture individuals likelihood to engage in, or avoid, creative behavior, and there is a strong case to be made that these differences in goal orientations lead individuals to respond differently to higher and lower levels of team bureaucracy. Second, bureaucracy can be characterized in terms of two core dimensions, centralization and formalization (Bolin & Ha renstam, 2008; Caruana, Morris, & Vella, 1998; Raub, 2007), and we propose that high versus low levels of centralization have different influences on individuals creative tendencies than high versus low levels of formalization. Our core contribution lies in an important advancement of person-in-situation analyses of creativity (Hirst et al., 2009; Taggar, 2002), demonstrating that contextual influences may not only invite the creative expression of individual differences, but also constrain them. Our study thus provides insights into resolution of the tension between bureaucratic control and innovation, which has eluded the field for half a century (cf. Burns & Stalker, 1961). This is also a pragmatically important issue, as organizations rely on a certain level of bureaucracy, prioritizing establishing and sticking to a beaten track, while also desiring creativity which by definition entails stepping off the beaten track. In examining person-in-situation influences, we deviate from the organizational design literatures study of the main effects of bureaucracy (e.g., Bunderson & Boumgarden, 2010; Raub, 2007) and bring together two perspectives on behavioral regulation that have been studied in separate traditions. One perspective has been characterized by individual-level analyses of self-regulation in goaldirected behavior (DeShon & Gillespie, 2005; Porath & Bateman, 2006); the other, by analyses of contextual regulation of behavior at the level of social aggregates (Raub, 2007). THEORETICAL BACKGROUND AND HYPOTHESES Individual creativity at work involves the development of practical and new solutions to workplace challenges, providing a tangible and useful outcome for an organization (Amabile, 1988). Following from the description of creativity as an outcome that derives from addressing work challenges, it is not surprising that individual differences in goal orientation that relate to an individuals motivation to tackle challenging problems

influence employee creativity (Gong, Huang, & Farh, 2009; Hirst et al., 2009). We first outline this individual difference perspective on creativity and then introduce the bureaucratic team context perspective, before we move on to our cross-level integration of these two perspectives in a series of hypotheses. Individual Team Members: Goal Orientations and Creativity Achievement motivation theory describes goal orientations as motivational orientations that capture how individuals regulate attention and effort when approaching, interpreting, and responding to achievement situations (DeShon & Gillespie, 2005; Elliot & Church, 1997). Two distinct orientations are commonly identified. A learning goal orientation (from here on, a learning orientation) is focused on the development of competence and task mastery and fosters an intrinsic interest in a task itself (Dweck, 1999). Intrinsic task motivation encourages individuals to invest effort and show perseverance (Amabile, 1996), and it is not surprising that a learning orientation encourages people to develop creative solutions to problems at work (Gong et al., 2009; Hirst et al., 2009; cf. Janssen & Van Yperen, 2004). People may also be motivated by extrinsic factors such as competing against others, receiving rewards, acknowledgement, or avoiding criticism (VandeWalle, 1997). This motivation is captured by the performance goal orientation, which is focused on the demonstration of competence to others. This externally attuned motivation can be divided into two subdimensions. A performance-prove goal orientation (from here on, a prove orientation) encourages individuals to seek to attain favorable judgments, whereas people who are concerned about avoiding unfavorable competence judgments have a performance-avoid goal orientation (from here on, an avoid orientation). The prove orientation may dispose individuals to be more creative when creativity is valued as a way to demonstrate competence (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 discourages these individuals from engaging in risky or challenging activities (VandeWalle, 1997) that would have provided opportunities for creativity. Goal orientation is fundamentally about self-regulation of behavior (Button, Mathieu, & Zajac, 1996; DeShon & Gillespie, 2005). Goal orientations encourage people to choose, either consciously or subconsciously, to engage in certain types of behaviors in achievement situations. For instance, in such situations, individuals with high levels of

626

Academy of Management Journal

June

learning orientation may choose to engage in adaptive behaviors patterns such as selecting challenging tasks, setting difficult goals, and persisting when obstacles are encountered. Therefore, to adequately model the behavioral outcomes of goal orientations, it is necessary to consider how they inform responses to the context in which the behavior is enacted (DeShon & Gillespie, 2005). More broadly, the strong emergence of person-in-situation approaches (Chen & Kanfer, 2006; Tett & Burnett, 2003; cf. Kristoff, 1996) highlights that an analysis of the interplay between individual and context is essential to predict the expression of individual dispositions. That is, the concept of disposition should not be misrepresented to imply that an individual will always behave in certain ways. Instead, a person-in-situation approach suggests moving away from a main effects approach in which the influence of individual differences is assumed to be constant. Rather, it implies an emphasis on the contingencies of disposition-outcome relationships. Individual creativity is often enacted in the context of a team or work group (Taggar, 2002), and this context may influence the relationship between goal orientation and creativity. Hirst et al. (2009) studied this very issue, focusing on team learning behavior as a contextual influence stimulating the expression of goal orientations that are conducive to creativity. They demonstrated that team learning behavior helped bring out the positive relationship between a learning orientation and creativity and between a prove orientation and creativity. Further testifying to the viability of this contingency perspective, Hirst et al. (2009) did not find relationships between goal orientations and creativity across the board but rather, found that the relationships were contingent on team learning behavior. These findings show that team dynamics may stimulate the expression of creative tendencies, yet they are mute on the issue that assumes center stage in the current study: the possibility that team contextual influences may also constrain the expression of creative tendencies. The absence of creativity-stimulating team contextual influences such as team learning behavior in no way imposes constraints on individuals creative behavior. Team bureaucratic practices, in contrast, represent a different class of team contextual influences in that they may impose exactly such constraints on creativity, and their influence cannot be extrapolated from earlier findings concerning creativity-stimulating influences (e.g., the absence of bureaucracy is not necessarily stimulating, just as the absence of creativity-simulating influences is not necessarily restraining). Adding to the complexity of the issue, the restraining influ-

ence of team bureaucratic practices may not have negative creativity consequences across the board, as we argue in the following. The Team Bureaucratic Context: Centralization and Formalization Teams, departments, and organizations differ in the extent to which bureaucratic practices restrain their members. Conceptual frameworks distinguish two main dimensions in this respect (Bolin & Ha renstam, 2008; Caruana et al., 1998; Raub, 2007; cf. Burns & Stalker, 1961): centralization of decision making (Van de Ven & Ferry, 1980) and formalization of rules and procedures prescribing and controlling behavior (Hall, 1999). Both centralized decision making and formal rules and procedures are ways of regulating and controlling employee behaviorthe essence of bureaucracyand are associated with low employee discretion on the job. Centralization relates to how power is distributed in an organizational hierarchy and whether employees are encouraged to participate in decision making (Hage & Aiken, 1967). Low centralization captures a context in which all employees participate and are afforded discretion and opportunities to act according to their own inclination. If decisions must be referred up the chain of command and made by a few superiors, centralization is high. Formalization relates to the extent to which rules are clearly specified and procedures standardized. Increasing formalization reduces the extent of employees freedom by prescribing procedures and potentially by sanctioning some courses of actionproviding specific directions as to appropriate actions, directing and enforcing these actions, and constraining employees ability to engage in discretionary behaviors (Raub, 2007). In the team context, centralization captures the extent to which within-team decision authority lies solely with a teams leader (decision making is centralized) or is shared between leader and members (decision making is decentralized and participative). Low centralization thus reflects an active influence on team members, who are expected to share decision-making authority with their leader. In that sense, low centralization may in fact have an influence that is described as empowering in other literatures in which decentralized, participative decision making is accorded an important role in actively engaging and intrinsically motivating employees (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 and procedures regulating team member behavior and

2011

Hirst, van Knippenberg, Chen, and Sacramento

627

thus if anything is merely a passive influence on individuals. Although high centralization and formalization may thus be similarly restraining in regulating and controlling team member behavior, we propose that low levels of centralization and formalization may in fact reflect markedly different influences on employeesan issue that becomes apparent when one considers their cross-level interaction with goal orientations. The importance of a focus on team bureaucracy as an influence on goal orientation-creativity relationships is evident in a conceptual analysis by Thompson (1965), who suggested that bureaucratic practices demotivate and restrain engagement in entrepreneurial, nonroutine tasks and limit innovation (cf. Raub, 2007). Such an inhibiting influence directly relates to the issues at stake in the current analysis. Situational factors that constrain or prescribe behavior limit the scope for the expression of individual differences by creating strong situations that override dispositions (Mischel, 1977). In contrast, situational influences may also invite (activate [Tett & Burnett, 2003]) the expression of individual dispositions. We propose that team bureaucratic practices are particularly relevant in this respect, because the level of team bureaucracy may both constrain and invite the expression of individual differences in goal orientations. In the following sections, we outline these propositions in more detail for both centralization and formalization. Individual Differences in a Team Context: Goal Orientation and Centralization In teams, centralization relates to the extent to which the leaders control and independently make decisions concerning the teams as opposed to engendering participative decision making (Bolin & Ha renstam, 2008). Centralized decision making reduces opportunities for individuals to contribute original thoughts or idiosyncratic novel views and to participate in discussions about important issues facing a teamthe very issues that might benefit from individual creativity in meeting challenges and solving problems. In contrast, decentralized decision making encourages individuals to contribute and participate in decisions and provides individuals latitude to express their views (Arnold et al., 2000; Chen et al., 2007). In decentralized contexts, team members share their views, which not only promotes a greater variety of different views, but may also invite a more committed and proactive engagement with potential challenges facing a team. The greater autonomy afforded by decentralization also fosters psychological ownership and empowerment and thus builds enthusiasm, com-

mitment to decisions, and intrinsic motivation, a powerful seed for creativity (Amabile, 1996). Decentralized practices promote a supportive climate that conveys an invitation to voice and share ones own perspective (Arnold et al., 2000) and therefore are also likely to reassure and encourage a teams members to feel that their contributions are appreciated (Wagner, 1994). This context may stimulate employees to grapple with less routine challenges and problems facing their team (Ahearne, Mathieu, & Rapp, 2005; Srivastava, Bartol, & Locke, 2006). Decentralized decision making thus may play an important role in bringing out the tendencies to engage in the creative behavior captured by learning and prove orientations and in attenuating tendencies to walk away from creative challenges associated with an avoid orientation. Learning-oriented people are motivated to learn and seek out tasks that are challenging because these provide greater opportunities for development. Centralization reduces opportunities to explore and learn and diminishes employees control. Centralization thus attenuates or removes the creative benefits of a learning orientation that derive from intrinsically motivated engagement with work challenges. Put differently, centralization inhibits (cf. Tett & Burnett, 2003) the creative expression of a learning orientation. Decentralized decision-making practices, in contrast, should appeal to learning-oriented individuals, activating (Tett & Burnett, 2003) their learning orientation, because of the opportunity to explore and learn from different views and ideas and because of the challenge and learning opportunities implicit in engaging with complex problems facing their team. This context may even create a positive cycle of interest and enthusiasm, encouraging higher intrinsic motivation that further fuels creativity (Amabile, 1996). Thus: Hypothesis 1a. A learning orientation and centralization interact to influence creativity: the learning orientation is positively related to creativity when centralization is low, but not when centralization is high. Performance-oriented people seek to maximize rewards and minimize possible punishments using environmental cues to decide which behaviors are appropriate (Seijts, Latham, Tasa, & Latham, 2004). They can be expected to focus on the cues, in this case the amount of centralization evident in a teams context, as a source of information for determining the choice of actions favored in the team. The nature of the influence of decentralized decision making on individuals with prove orientations compared to those with avoid orientations will differ, however.

628

Academy of Management Journal

June

For prove-oriented people, motivation to engage in creative activities depends on the extent to which they perceive they are likely to be recognized and rewarded for their behavior. Accordingly, we may expect individuals with a stronger prove orientation to be more responsive to the extent to which their context encourages them to contribute to team decisions and to provide suggestions to solve team problems. When prove-oriented individuals work in a context where the leader arrives at decisions with little consultation or invitation to participate, they may assume the leader has little interest in their suggestions or views and become less likely to engage in complex problem solving and creativity to meet work challenges; centralization inhibits the creative expression of a prove orientation. In contrast, in decentralized contexts, where team members are encouraged to contribute, prove-oriented people will construe team discussions as a forum in which to demonstrate competence. The chance to display their prowess may tantalize them to display high levels of proficiency and to be acknowledged and recognized for their abilities. For prove-oriented people too, decentralization works as trait activating. We therefore expect that a prove orientation is related to creativity when centralization is low rather than high. Hypothesis 1b. A prove orientation and centralization interact to influence creativity; the prove orientation is positively related to creativity when centralization is low, but not when centralization is high. For an avoid orientation and centralization, matters are different. Unlike learning and prove orientations, which may inspire active engagement with job challenges, avoid orientations predispose individuals to stay clear of job challenges and problems that may invite acts of creativity. As a result, the lack of support and appreciation for employees proactive engagement with the work that is conveyed by high centralization may actually reinforce avoid-oriented individuals dispositional tendencies, whereas the active, engaging influence of decentralized decision making may work to counter avoidant tendencies. For individuals with an avoid orientation, centralization creates a team climate with limited support and few opportunities to proactively contribute to decisions and solve work challenges that require creativity. In these settings, absence of explicit encouragement to creatively engage with important job issues may lead avoidoriented people to perceive their leader as less than supportive of experimental and potentially risky activities, reinforcing their tendency to avoid such activities, which would have provided greater oppor-

tunities for creativity. Thus, in contrast to its inhibiting influence on the relationships between learning and prove orientations and creativity, if anything centralization may bolster the negative relationship between an avoid orientation and creativity. A different set of events occurs in decentralized contexts, where participative practices signal that individuals contributions and proactive engagement with work challenges are not only supported and encouraged, but also appreciated and expected. This setting fosters a climate in which norms support participating in decisions and engaging with work challenges. This climate may stimulate or at least reassure avoid-oriented individuals that their proactive engagement with work challenges is appreciated, even though these behaviors may differ from their inherently cautious inclinations. This sense should lower their dispositional barriers to trying new approaches, making it easier for avoid-oriented individuals to clarify uncertainties by facilitating shared learning through team discussion and problem solving. Likewise, decentralized practices that foster inclusive decision making diminish the negative consequences of voicing an opinion that may be out-ofsynch with the views of ones team, thus lowering the threshold for engaging in such behaviors, because it signals both that false assumptions or errors are less likely to elicit negative feedback and that team members are encouraged to discuss different approaches. Thus, we predict: Hypothesis 1c. An avoid orientation and centralization interact to influence creativity; the avoid orientation is negatively related to creativity when centralization is high, but not when centralization is low. Individual Differences in Team Context: Goal Orientation and Team Formalization Formalized team contexts are characterized by rules and standardized procedures that limit the choice of behaviors and decisions an individual can make. Rules and regulations effectively reduce individuals opportunities to engage in discretionary behaviors and provide a setting that presses individuals to follow certain procedures and approaches (Hall, 1999). Teams that are high in formalization are characterized by clear behavioral protocols involving administrative checks (as evidenced by paperwork and administration) that regulate and direct employees behavior. Thus, they are prototypical instances of what Mischel (1977) called strong situations. As formalization increases, behavior is regulated and restrained to a

2011

Hirst, van Knippenberg, Chen, and Sacramento

629

greater degree and funneled and homogenized to a small set of choices offering little freedom, reducing the expression of individual differences. In terms of trait activation theory (Tett & Burnett, 2003), high formalization inhibits the expression of individual dispositions. In less formalized team contexts, there will be fewer guidelines, more opportunities for discretion, and thus more leeway for the expression of individual differences. This is not to say, however, that the absence of formalized practices encourages creativity (i.e., unlike decentralized decision making, which plays a more active role in this respect). Rather, it points to greater freedom for the expression of individual differences with fewer specific procedural directions to follow, which may either stimulate or impair creativity, according to an individuals disposition. Learning-oriented employees are intrinsically motivated to learn. A highly formalized team environment restricts the expression of this desire by limiting discretion, thus attenuating the influence of a learning orientation on creativity. In comparison, learning-oriented employees working in less rule bound team environments experience fewer restrictions and have greater discretion to express their inclination to try new approaches. Accordingly, as a learning orientation disposes individuals to creatively engage with work problems and challenges, we predict: Hypothesis 2a. A learning orientation and formalization interact to influence creativity; the learning orientation is positively related to creativity when formalization is low, but not when formalization is high. Performance-oriented individuals in teams with high formalization perceive clear guidelines as to the behavior that is expected and seen as appropriate. Prove-oriented individuals use these guidelines to regulate their behavior, which results in rule following and unwillingness to engage in behaviors deviating from these standard procedures. In effect, this reduces prove-oriented individuals creative tendencies. Teams with little formalization of rules and procedures, on the other hand, provide an environment in which there is less enforced clarity as to desired ways to engage with the job. Prove-oriented individuals seeking to demonstrate high performance in a context with few formal rules and procedures thus have little information as to how they can demonstrate competence in relation to their peers other than by getting the job done, and done well. Compared to situations of high formalization, this situation may invite more creativity in meeting work challenges from individuals with a stronger motivation to demonstrate

their competence. Developing creative solutions to problems others have failed to resolve provides a particularly powerful demonstration of competence and so provides an opportunistic chance to demonstrate ones capability, encouraging proveoriented employees to engage in creative problemsolving behaviors. Hypothesis 2b. A prove orientation and formalization interact to influence creativity; the prove orientation is positively related to creativity when formalization is low, but not when formalization is high. For an avoid orientation, too, there is reason to expect a less evident influence in teams with higher formalization of practices. The reason for this is the same as it is for learning and prove orientations: The more rules and procedures guide job performance, the less freedom there is for the expression of individual differences. In teams low in formalization, however, the absence of prescribed structure and the associated freedom may increase uncertainty about the appropriate ways to engage with the job. Avoid-oriented individuals tend to avoid challenges that carry the risk of errors and failure and to favor endeavors with a high chance of success (VandeWalle, 1997). When guidelines are few, the risk of failure may loom large and invite avoid-oriented individuals to steer clear of creative challenges. Thus, whereas formalized procedures may do little to promote creativity, they at least work to keep avoid-oriented individuals tendency to avoid creative challenges in check, rendering the relationship between an avoid orientation and creativity weaker under conditions of high formalization than under conditions of low formalization. Note that this is where the difference between low centralization and low formalization discussed earlier expresses itself. Whereas decentralization in fact is an active influence that makes the avoid orientation creativity relationship weaker than it is under conditions of high centralization (Hypothesis 1c), low formalization merely reflects the absence of rules and procedures regulating behavior and actually provides more opportunity for the expression of an avoid orientation than high formalization. Thus, we hypothesize: Hypothesis 2c. An avoid orientation and formalization interact to influence creativity; the avoid orientation is negatively related to creativity when formalization is low, but not when formalization is high.

630

Academy of Management Journal

June

METHODS Research Setting, Participants, and Procedures A national survey of employees of the Taiwan Customs Bureau was conducted, comprising those working at the Bureaus offices at the international airports of Taipei and Kaohsiung and the international harbors of Keelung, Taichung, and Kaohsiung. Employees worked in teams that were responsible for a particular activity (e.g., developing finance protocols or search procedures), worked relatively independently of each other, and had leaders with a high degree of influence on teamwork outcomes. Forty-three percent of the teams performed operational search and detection tasks. Given the strategic sensitivity of the region and the need to keep pace with a technology and crime arms race (Ball, 2003) in which the nature and methods of offenses varied (from trafficking controlled substances, narcotics, animal and farm produce, protected species, and organs and biomedical by-products, to visa fraud and people smuggling), this was a particularly challenging function. Creative problem solving was required to develop new innovative detection approaches to keep pace with diverse criminal activities. For example, creativity was required to teach scent-detecting dogs trained in another country to identify new scents associated with offenses specific to the Taiwanese context (e.g., betel nut smuggling). The remaining 57 percent of the teams performed administrative and management functions such as performance monitoring, information technology, finance, and human resources activities. Given a strategic mandate to work and share data with other agencies, the Customs Bureau placed much emphasis on systems development and integration. For example, since the September 11, 2001, attacks, Taiwan Customs have worked with U.S. government agencies to share security information, and this interaction has necessitated new computer systems as well as the development and application of new X-ray inspection technologies for cargo and containers. Thus, a combination of challenging demands and limited resources made it essential for employees to display creativity in developing innovative solutions to problems. Recognizing the necessity of creativity and innovation, management implemented a department-wide total quality management (TQM) initiative that included appraisals as well as monetary rewards or credits to encourage employees to submit creative ideas to improve work processes or solve work problems. With these approaches, the Bureau sought to empower officers to streamline work practices to reduce red tape, yet also assure they worked within government regulations and demonstrated procedural compliance.

Employees creativity was appraised in annual performance reviews and promotion decisions. Performance reviews included criteria assessing how subordinates used creativity in their jobs (e.g., by developing processes to improve work efficiency). Promotion decisions also assessed the creativity of employees researching and working. For example, a subordinate grappling with a new data-collecting system would be appraised on whether he or she had developed new ways to address or solve problems. Subordinates who achieved positive assessments obtained credits in their promotion evaluation. The survey was translated by one of the authors and two university faculty using the procedures described by Brislin (1980). At the projects initiation, we conducted a small pilot test with 36 employees to make sure all measures were reliable and used terms appropriate to the organization. Having obtained adequate to good psychometrics and verbal feedback from participants, we slightly refined the translation of some of the measures for the larger study. A rank-and-file, long-standing member of the organization who was supportive of the broader project personally distributed a paper survey to teams, ensuring a 100 percent completion rate. In total, 388 employees completed questionnaires containing the independent variables, and 97 team leaders rated employee creativity. As in previous 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 evaluate their employees and to be comfortable, willing, and fairly adept at doing so. Prior research supports the validity of this measurement approach. Supervisor ratings have been found to relate to creative output (e.g., suggestions for improvement schemes, invention disclosure forms and patents [Tierney, Farmer, & Graen, 1999]). Participants had the option of responding anonymously, which precluded matching questionnaires to supervisor ratings, and in addition some records were obsolete, either because employees had left the organization or administrative records were incomplete. We therefore obtained matched data for 330 employees and 95 teams, providing a usable sample of 90 and 97 percent of the original individual- and team-level populations. The majority of sample members were male (73%); the average age was 44 years, and average tenures in the organization and in their current position were 16 and 6 years. A quarter of the sample had been educated at junior college (the equivalent of a short, practically oriented university program); 64 percent had university qualifications; and 10 percent had postgraduate degrees. Em-

2011

Hirst, van Knippenberg, Chen, and Sacramento

631

ployees were paid at varying salary bands according to their length of service. More senior employees who were on higher bands tended to work in roles that provided greater decision-making latitude. Measures Goal orientation. We used VandeWalles (1997) three-factor scales to assess goal orientation as learning, avoid, or prove. The rating scale for all items ranged from 1 not at all, to 7 to a large extent. An example of the five-item learning orientation scale is I often look for opportunities to develop new skills and knowledge. An example of the four-item avoid orientation scale is Im concerned about taking on a task at work if my performance would reveal that I had low ability. An example of the four-item prove orientation scale is Im concerned with showing that I can perform better than my co-workers. Centralization. In selecting a centralization measure, we also reviewed scales constructed for an organizational context rather than a team context (e.g., Lee & Choi, 2003; Raub, 2007). Because team context was a core element in both our theory development and research setting, however, we were left with the choice of adapting one of these scales or selecting a team-focused alternative. Given the good reliability and validity reported for it, we selected Arnold et al.s (2000) four-item measure, which assesses the extent to which within-team decision making is centralized in the leader role as opposed to decentralized in participative decision making. The items, rated from 1, strongly disagree, to 7, strongly agree, assess centralization of decision making, using a teams leader as the referent, as follows: Uses my work groups suggestions to make decisions that affect us, Listens to my work groups ideas and suggestions, Encourages work group members to express ideas/suggestions, and Gives all work group members a chance to voice their opinions. To facilitate interpretation, we reverse-scored the scale so that higher ratings reflected greater centralization of within-team decision making. Formalization. We used a three-item scale (Hage & Aiken, 1967; Rafferty & Griffin, 2004) relating to the extent to which work was structured and regulated 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 this team, Our work involves a great deal of paperwork and administration, and Our work is highly regulated by bureaucratic procedures. Creativity. We measured employees creativity using the four-item scale reported by Farmer, Tier-

ney, and Kung-McIntyre (2003). Team leaders rated employees creativity on a scale ranging from 1, not at all correct, to 6, completely correct. Items were Seeks new ideas and ways to solve problems, Generates ideas revolutionary to the field, Is a good role model for innovation/creativity, and Tries new ideas and approaches to problems. Control variables. We controlled for gender (0 male, 1 female), education (1, high school, to 5 postgraduate degree: masters or Ph.D.), individual age and mean team age, individual and team tenure, and team size, as each has been found to relate to employee creativity (Tierney & Farmer, 2002) and to influence team processes (Hirst et al., 2009). We included additional controls to take into account both the heterogeneity of the sample and practices particular to the Customs Bureau. We controlled for employees remuneration band (i.e., more senior employees were generally paid more and afforded greater discretion and so greater opportunities for creativity). At the team level, we controlled for whether a team was operational (coded 1) or administrative/managerial (coded 2). To control for potential differences in offices and practices, we created two dummy variables for region: region 1 was Taipai, the capital and largest airport, and region 2 was Keelung, a distinctive locale that hosts the most remote provincial office, which is responsible for large volumes of sea-based trade through Keelung Harbor. Validation of Multilevel Data Structure Analyses included individual- and team-level constructs. Performing two different analyses to validate this data structure, we first examined whether the data justified aggregation of team-level constructs. According to a one-way analysis of variance (ANOVA), centralization and formalization differed between teams (p .05). Mean values of rwg(j) across teams of .89 for centralization and .92 for formalization, calculated using a normal distribution, suggested adequate within-team agreement (James, Demaree, & Wolf, 1984). Thus the rwg(j) supported aggregation of constructs to the team level. For a more comprehensive test of the multilevel data structure, we conducted a multilevel factor analysis following procedures recommended by Dyer, Hanges, and Hall (2005). We first performed confirmatory factor analysis (CFA) on the six constructs of learning, prove, and avoid orientations, and centralization, formalization, and creativity. The fit of the model with six factors loading separately (2 493.26, df 237, p .01, RMSEA .06, CFI .95) was compared with the fits of six potential alternatives. The first alternative was a

632

Academy of Management Journal

June

five-factor model with prove and avoid orientations collapsed to represent an overall performance orientation factor (2 818.06, df 242, p .01, RMSEA .09, CFI .88), and the second alternative model had the learning and prove orientations collapsed into one approach factor (2 1,224.97, df 242, p .01, RMSEA .11, CFI .80). The third alternative was a four-factor model collapsing all three orientations into one goal orientation factor (2 1,565.06, df 246, p .01, RMSEA .13, CFI .73); next, we tested a threefactor model also collapsing centralization and formalization into one bureaucracy factor (2 1,664.81, df 249, p .01, RMSEA .13, CFI .71); a two-factor model collapsing all independent variables into one factor (2 2959.09, df 251, p .01, RMSEA .18, CFI .44); and a one-factor model (2 3,458.06, df 252, p .01, RMSEA .20, CFI .33). These findings show that the sixfactor model provided a good fit with the data and a better fit than the alternative models. These data modeled at the individual level thus demonstrated the convergent and discriminant validity of the constructs studied, providing sufficient basis to test the multilevel structure of the data (Dyer et al., 2005). We expected that the sixfactor structure of the model would be consistent at both levels; thus we constructed within- and between-team CFA models comprising six factors. Multilevel factor analysis is computationally demanding, and convergence problems tend to arise when the number of observed indicators is high (Dyer et al., 2005). To avoid nonconvergence, we used item parcelling based on item skewness to reduce the number of observed indicators (Nasser & Wisenbaker, 2003). This procedure resulted in a reduction in the number of indicators from 24 to 13 and a factor structure with acceptable fit (Muthe n, 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 aggregated variables above the diagonal. Our hypotheses are cross-level interaction hypotheses, and we used hierarchical linear modeling (HLM) to test them. Following Preacher, Curran, and Bauers (2006) recommendations, we group-mean-centered all individuallevel (level 1) variables except for gender. Team-level (level 2) variables were not centered, to reduce possible 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 individuallevel variables (step 1), followed by the team variables (step 2), and in the final step we simultaneously tested all cross-level interactions (step 3). The null model allowed us to test the betweenteam variance in creativity by examining the level 2 residual variance of the intercept (00) and the ICC1 statistic, which represents the proportion of variance in the outcome variable that resided between groups. Analyses revealed a 00 of .14 (p .001) and an ICC1 of .37, indicating 37 percent of the variance resided between teams, to be explained by level 2 variables. A precondition for testing crosslevel interactions here was that the slopes of relations between the goal orientations and creativity vary across teams. This was the case (learning orientation: 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 of Table 2 show the level 1 and 2 main effects. Of the level 1 control variables, only age was a significant, and negative, predictor of creativity ( .03, p .05). Avoid orientation ( .08, p .05) had a negative relation with creativity. None of the level 2 variables were significantly related to employee creativity. Note that although the graphs for the interactions presented below might suggest main effects for the dimensions of bureaucracy, these were not significant. These figures should be interpreted in light of the fact that we followed Aiken and Wests (1991) method for examining interactions, whereby values are selected by convention (i.e., one standard deviation above and below the mean). Were other values to be selected, the results might differ, and so these figures should be interpreted in terms of the slopes for the goal orientations, and not the intercept for low and high bureaucracy. Moreover, we targeted these graphs to reflect one particular moderating influence on the relationship of one particular goal orientation with creativity and did not fully factor in the influence of the other model variables. Cross-level interactions. We estimated slopes-asoutcomes models in HLM to assess the moderating effect of centralization and formalization on the relationship between goal orientation and employee creativity. The final section of Table 2 presents the results of this analysis. The model explained 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 crosslevel interactions. Hypotheses 1a, 1b, and 1c predict that centralization moderates the relations be-

TABLE 1 Means, Standard Deviations, and Correlationsa

1 .23 .89** .63** .32** .04 .14** .17** .12* .19** .09 .34** .53** .03 .02 .10 .12* .32** .28** .09 .07 .20** .10 .20** .10 .18** .26** .38** .08 .06 .06 .02 .10 .00 .36** .02 .40** .00 .09 .09 .05 .06 (.89) .04 .09 .02 .47** .11 .07 .14* .07 .07 .01 .20** .04 .29** .01 .01 .03 .02 .19** .01 .15** .09 .11* .03 .11 .11 .15 .03 .20* (.82) .47** .03 .00 .02 .03 .03 .06 .02 .04 .13* .34** .17 .09 .12 .18 .01 .13 .07 .28** .21 .06 .24** .87** .62** .62** .66** .35** .06 .03 .10 .09 .14** .14** .18** .05 .08 .24** .17** .06 .08 .32** .21** .63** .39** .04 .03 .06 .06 .04 .01 .08 .02 .07 .01 .07 .11 .10 .44** .22* (.93) .07 .08 .07 .02 .05 .04 .11* .11* .04 .42** .15 .03 .12 .09 .10 .02 .15 .04 .27** .32** .28** .14 .14 .15 .02 .32** 2 3 4 5 6 7 8 9 10 11 12 .23** 1.00** .87** .62** .36* .02 .11 .01 .15 .27** .17 13 .06 .87** 1.00** .62** .40** .01 .11 .01 .03 .32** .09 .87** .17** .09 .52** .04 .08 .06 .08 .10 .04 .01 .07 .05 .08 .08 .12* 14 .18 .09 .04 .05 .01 .06 .07 .05 .21* .17 .04 .09 .04 15 .08 .11 .12 .11 .05 .14 .07 .12 .13 .05 .15 .11 .12 .04 .04 (.99) .04 .01 .13* .04 16 .14 .01 .06 .06 .04 .12 .01 .02 .24* .09 .01 .01 .06 .13 .03 (.65) .02 17 .00 .11 .11 .04 .05 .04 .15 .11 .15 .03 .12 .11 .11 .17 .05 .03 (.84)

Variable

Mean

s.d.

1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17.

Gender Age Tenure Revenue Education Learning orientation Avoid orientation Prove orientation Category Region 1 Region 2 Mean age Team tenure Team size Centralization Formalization Creativity

0.31 0.47 44.01 8.05 .21** 16.14 9.17 .06 82.39 14.60 .18** 2.81 0.63 .03 4.79 1.05 .02 4.12 1.12 .13* 4.10 1.22 .14* 1.43 0.50 .26** 0.29 0.46 .01 0.07 0.26 .19** 43.99 4.83 .08 15.81 5.91 .00 6.98 2.01 .14* 2.79 1.07 .03 3.70 0.62 .06 3.61 0.60 .01

a Individual-level analyses are below the diagonal, and group-level analyses are above the diagonal. Internal consistency reliabilities are in parentheses. n 95 teams comprising 330 employees. * p .05 ** p .01

634

Academy of Management Journal

June

TABLE 2 HLM Results for the Effects of Cross-Level Interactions of Goal Orientation with Team Bureaucracy on Employee Creativity
Variable Null model Intercept Level 1 variables Intercept Gender Age Tenure Revenue Education Learning Prove Avoid Level 2 main effects Category Region 1 (two offices) Region 2 (four offices) Mean age Team tenure Team size Centralization Formalization All interaction terms Learning orientation centralization Prove orientation centralization Avoid orientation centralization Learning orientation formalization Prove orientation formalization Avoid orientation formalization
a

Coefficient

s.e.

Model Deviance

R2b

2c Rtotal

3.62* 3.63* 0.09 0.03* 0.01 0.01 0.03 0.05 0.03 0.08* 0.07 0.01 0.14 0.01 0.02 0.04 0.02 0.01

.05 .05 .08 .01 .01 .00 .06 .03 .03 .03

77.09* 78.13* 1.24 2.47* 1.37 1.70 0.50 1.49 0.95 2.28* 0.63 0.10 0.82 0.61 1.74 1.66 0.35 0.19

285.33 304.64

585.15 584.49 .02

.12 .12 .18 .01 .01 .02 .04 .05

281.20

612.31

.00

.02

0.06* 0.03 0.08* 0.02 0.10* 0.08*

.03 .03 .03 .04 .04 .03

1.97* 1.11 2.58* 0.59 2.70* 2.26*

297.92

623.12

.03

.04

Employee n 330, team n 95. 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 R2total R2within-group (I ICC1) R2between-groups ICC1. * p .05 Two-tailed tests.
b

tween creativity and the learning, prove, and avoid orientations. The interaction of learning orientation and centralization was significant ( .06, p .05). Figure 1A depicts this interaction. The slope was significant and positive when centralization was low (one standard deviation below the mean; .12, t 3.20, p .01) and nonsignificant when it was high (one standard deviation above the mean; .01, t .19, n.s.). These results support Hypothesis 1a.1 The prove orientation by central1

ization interaction was not significant, failing to support Hypothesis 1b. The interaction of the avoid orientation and centralization was significant ( .08, p .05; see Figure 1B). The slope was significant and negative when centralization was high ( .16, t 3.74, p .01) and nonsignificant when centralization was low ( .01, t .27, n.s.). These results support Hypothesis 1c. Hypotheses 2a, 2b, and 2c predict that formalization moderates the relation between creativity and learning, prove, and avoid orientations. Not

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

both team-level variables. For exploratory purposes and in anticipation of readers potential interest, we also tested three-way interactions reflecting possible interactive effects of centralization and formalization. None of these interactions were significant.

2011

Hirst, van Knippenberg, Chen, and Sacramento

635

FIGURE 1 Interactions for Centralization

(1A) Interaction of Centralization and Learning Orientation Predicting Creativity 5

4.5
Employee Creativity
High centralization Low centralization

3.5
Low High

Learning Orientation (1B) Interaction of Centralization and Avoid Orientation Predicting Creativity
5

4.5

Employee Creativity
4

3.5
Low High

Avoid Orientation

supporting Hypothesis 2a, the learning orientation by formalization interaction was not significant. The interaction of a prove orientation and formalization was significant ( .10, p .05; see Figure 2A). The slope was positive and significant when formalization was low ( .11, t 2.87, p .01) and negative and nonsignificant when it was high ( .06, t 1.03, n.s.). These results support Hypothesis 2b. The interaction of the avoid orientation and formalization was significant ( .08, p .05; see Figure 2B). The

slope was significant and negative when formalization was low ( .14, t 3.42, p .01) and nonsignificant when it was high ( .00, t 0.03, n.s.). These results support Hypothesis 2c. DISCUSSION Managers face the challenge of getting creative results from the individuals in their teams, yet a desire for control and behavioral regulation may inspire centralized decision making and formal-

636

Academy of Management Journal

June

FIGURE 2 Interactions for Formalization

(2A) Interaction of Formalization and Prove Orientation Predicting Creativity 5

4.5

Employee Creativity
4

High formalization Low formalization

3.5
Low High

Prove Orientation

(2B) Interaction of Formalization and Avoid Orientation Predicting Creativity 5

4.5

Employee Creativity
4

3.5
Low High

Avoid Orientation

ization of rules and procedures. Our findings confirm that team bureaucracy can suppress the expression of individual differences that may engender creativity. A learning orientation was positively related to creativity only under conditions of low centralization, and a prove orientation was positively related to creativity only when formalization was low. Moreover, low centralization attenuated the negative association between an avoid orientation and creativity. Although the evidence would seem to favor lower bureaucracy for engendering creativity, findings for the avoid orientation by formalization interaction provide an important caveat: less formalized practices were associated with a more negative relation-

ship between an avoid orientation and creativity. These findings illustrate that different goal orientations produce different creative responses to a given context and suggest the promise of personin-situation perspectives for understanding team bureaucratic influences. Theoretical Implications Although it is well established that both individual differences and context play an important role in the creative process (Amabile, 1996; Shalley et al., 2004), person-in-situation approaches that account for their interactive influences are still in development. Goal orientations are of particular

2011

Hirst, van Knippenberg, Chen, and Sacramento

637

interest in this respect, given their strong links with behavior in achievement situations and reviews highlighting the need to examine goal orientations as they unfold dynamically in context (Button et al., 1996; DeShon & Gillespie, 2005). In an important counterpoint to earlier work by Hirst et al. (2009) demonstrating that team contextual influences may stimulate the creative expression of goal orientations, the present study shows that the team context may inhibit as well as stimulate the link between goal orientations and creativityand, moreover, that this inhibition is not necessarily bad for creativity. These findings are important in suggesting that the influence of a situation on the creative expression of dispositions should be understood not only in terms of the extent to which the situation provides cues that may activate the expression of traits, but also in terms of the extent to which it constrains individuals behavior and thus inhibits the expression of individual differences. This argument suggests the contours of a more comprehensive model of team contextual influences on the relationship between individual differences and creativity, a model that incorporates both traitactivating and trait-inhibiting influences. Although such a model is clearly consistent with trait activation theory (Tett & Burnett, 2003), the obvious challenge for creativity research is to identify specific trait-context combinations that are relevant to creativity. An important issue to take into account in this respect is that the present findings for centralization show that the exact same contextual cue may activate one trait while inhibiting another. For research in creativity, the findings for team bureaucracy are important in that they show that the creativity of those motivated to engage with work challenges (i.e., learning- and prove-oriented individuals) benefits from low bureaucracy. It is less clear why, for learning-oriented people, this relationship holds for low centralization but not low formalization, whereas for prove-oriented individuals, it would hold for low formalization but not for low centralization. We are hesitant to base conclusions on null findings (i.e., the absence of learning orientation by formalization and prove orientation by centralization interactions). Consistently with predictions, the pattern of results is more complex for the avoid orientation and bureaucracy. Decentralized decision making helped attenuate the negative relationship between the avoid orientation and creativity, but low formalization actually brought out this relationship. As we outlined in the introduction, this can be seen as reflecting an important difference between the active and engaging nature of low centralization as compared with the passive nature of low formalization.

Our findings may also provide important pointers for identifying further ways to manage the undesirable outcomes of an avoid orientation (Dweck, 1999). Encouraging avoid-oriented individuals to participate makes it more likely that they will overcome their disposition to avoid job challenges. Findings for team bureaucracy also speak to the value of taking lower levels of analyses into account than is typically done in research on bureaucracy and related governance issues (Bolin & Ha renstam, 2008; Caruana et al., 1998; Hall, 1999). To reconcile the tension between bureaucracy and creativity, individual differences should be considered as important moderators of the effectiveness of organizational practices. In other words, these individual dispositions provide insight as to why different people respond so differently to the same bureaucratic experiences. This in turn sheds light on how to develop more constructive solutions to the tension between managerial control and innovation. We have pointed out that the available evidence for goal orientation creativity relationships suggests that these should be expected to manifest not as main effects, but rather, as contingent on contextual influences. The present findings by and large corroborate this analysis. Only the avoid orientation had a main effect; the learning and prove orientations did not. As the evidence for the contingent nature of the goal orientation creativity relationship amasses, it is interesting to note that main effects of goal orientations for in-role task performance have been more consistently documented (Porath & Bateman, 1996; VandeWalle, Brown, Cron, & Slocum, 1999). This could reflect fundamental differences between creativity and inrole performance that render creativity or at least the expression of individual differences in creativityinherently more contingent on contextual influences. Future research that combines the study of creativity and in-role performance may be highly worthwhile as a way to shed more light on this issue. The systematic development of the goal orientation literature has been hampered by differing conceptualizations of the construct and the wide array of scales in use; no common definition or set of measures of goal orientation exists (DeShon & Gillespie, 2005). Our framework (cf. Elliot, 1999; Elliot & Church, 1997) relies on a three-dimensional conceptualization (Seijts et al., 2004; VandeWalle, 1997) that partitions performance orientation into its prove and avoid components. Although this is the most widely accepted conceptualization of goal orientation, we recognize that an earlier two-factor conceptualization grouped the prove and avoid orientations under a general performance orientation construct (Button et al., 1996; Stevens & Gist, 1997).

638

Academy of Management Journal

June

In this respect, our findings clearly testify to the validity and value-added of the distinction between prove and avoid performance orientations. CFA showed that these represent two distinct factors, and relationships with creativity clearly diverge for prove and avoid orientations. We note that on similar grounds, our findings also speak against grouping learning and prove orientations under a more general approach motivation construct (cf. Elliot, 1999). CFA demonstrated that these are distinct factors, and relationships with creativity also diverged. The current three-dimensional goal orientation framework thus can be seen as providing the most valid basis for the interpretation of our results. Managerial Implications Bureaucratic practices often hinder managers efforts to facilitate individual initiative and creativity (Bolin & Ha renstam, 2008). To get creative results, managers should seek to understand employees motivational orientation in context, so the combination of individual disposition and bureaucratic context yields the most desirable associations with creativity. Decentralization brings out the best in learning-oriented employees and attenuates the negative effects of an avoid orientation on creativity. Thus, from a creativity perspective, decentralized decision making is attractive, even when it does little for prove-oriented individuals. Formalized practices are a more complex issue. Although the creativity of prove-oriented individuals thrives in less formalized contexts, low formalization actually brings out undesirable tendencies in avoid-oriented team members. Given its very nature, formalization cannot be tailored to individual team members, and these findings rather speak to the issue of selection and person-environment fit. An avoid orientation brings few creative benefits. At best, it is not negatively related to creativity, but achieving such an absence of association relies on formalizing procedures at the expense of the creative benefits of a prove orientation. Accordingly, the conclusion would be that creativity is best obtained by (where possible) not selecting avoid-oriented individuals for a team, while creating a team context characterized by high levels of decentralized decision making and low formalization, to let learning-oriented and prove-oriented individuals flourish. Limitations and Directions for Future Research We tested how individuals responded to a bureaucratic context but did not test the mediating

processes underpinning these relationships. In this respect, the goal orientation framework suggests that different goal orientations lead individuals to respond to different aspects of the same situation. Whereas learning-oriented individuals may respond to challenges and opportunities for learning and further development, prove-oriented individuals may respond primarily to cues that identify opportunities to display their competence. Avoidoriented individuals, in contrast, seem more driven by cues highlighting the safety or potential risks associated with engaging in certain actions in terms of how others will assess their performance. Future research tapping perceptions of situations along those lines as potentially mediating processes may further validate the current cross-level perspective on individual creativity. It may also provide important information for interpreting our less expected findings, such as the absences of a learning orientation by formalization interaction and of a prove orientation by centralization one. It is possible that situations of low centralization and low formalization, although both reflecting low bureaucracy, differ in the eyes of learning-oriented and prove-oriented individuals as to the key triggers of their motivational drives (e.g., for learning-oriented individuals, decentralization may be more associated than low formalization with learning opportunities). We examined antecedents of employee creativity and did not test whether these effects generalized to in-role performance (i.e., activities mandated in an employees position description and work designation). We highlight this is as a much-needed area for future work, as creativity and performance researchers have studied similar constructs but conducted few comparisons between the two. Research examining both outcomes would help explain whether different antecedents are required to stimulate them and thus would allow tests of the generalizability of different theoretical frameworks. We also suggest there is a need to understand how cultural values impinge on the utility of goal orientation theory, which is derived from European-American values. Results for East Asian samples, such as the correlations between learning orientation and creativity observed in the current study (r .11) and by Gong et al. (2009; r .20), are consistent with Harrison, Neff, Schwall, and Zhaos (2006) results in a meta-analysis of EuropeanAmerican research (range of r .09 .30; mean r .17). Yet one can speculate as to how cultural values may also influence the predicted relationships. Extending the current framework, we suggest that a culture such as Taiwan, characterized by higher uncertainty avoidance than the United States, may be more likely to invite the expression of an avoid orientation at the expense of creativity. That is, all other things

2011

Hirst, van Knippenberg, Chen, and Sacramento

639

being equal, the avoid orientation may have a stronger impact on creativity in, for instance, Taiwan than in the U.S. We also see parallels between centralization and power distance (Hofstede, Neuijen, Ohayv, & Sanders, 1990) that would suggest that, all else being equal, low power distance cultures are more likely to invite creativity based on a learning orientation. Similarly, the more collectivistic orientation of a country such as Taiwan as compared to, for instance, the U.S., may also play a role. A performance approach orientation has a strong connotation of individual competition, which may be more easily activated in an individualistic than a collectivistic culture. The more general point is that culture can be seen as context too, and as such cultural context may influence goal orientation creativity relationships. Expanding knowledge of different indigenous contexts and thus providing globally informed perspectives is important for management research if scholars are to speak to international business and commerce trendsparticularly as East Asian economies look set to overtake established European and American economies in the next decades. We note that effect sizes for the interactions were small. An important issue in this respect is that survey research can be assumed to lead to underestimating the effect size of interactions (Evans, 1985; McClelland & Judd, 1993). This not only means that it is difficult to detect interactions in survey research (cf. Harrison, Price, & Bell, 1998), but also that effect size may not be a fair criterion for judging them (Evans, 1985). Evans suggested that, given this underestimation problem, interactions with explained variance of as little as 1 percent should be taken seriously, as they most likely represent larger true effect sizes. In view of these considerations as well as more general evidence that goal orientations may not have simple direct relationships with creativity (e.g., Hirst et al., 2009), we would argue that establishing the current moderated relationships is important to understanding creativity at work. Even when explained variance appears to be small, it builds fundamental theory to understand the team contingencies relevant to the expression of individual differences in creativity, and such understanding represents an important step toward identifying the conditions under which goal orientations are most relevant to creativity. Conclusion Our study shows that team bureaucratic practices may regulate and influence individuals goal-directed behavior to activate as well as inhibit the creative expression of goal orientations. In this respect, decentralization delivers the greater creative benefits, invit-

ing creativity from those intrinsically inclined to learn and attenuating individuals avoidant tendencies. This is not to say that bureaucratic practices always diminish creativity; indeed, they may also curb less desirable avoidant tendencies. As bureaucratic practices do little to bring out creativity, the challenge is therefore to encourage creativity by reducing centralization and formalization, while not inviting individuals creativity-avoidant tendencies. REFERENCES
Adler, P. S. 1999. Building better bureaucracies. Academy of Management Executive, 13(4): 36 47. Ahearne, M., Mathieu, J., & Rapp, A. 2005. To empower or not to empower your sales force? An empirical examination of the influence of leadership empowerment behavior on customer satisfaction and performance. Journal of Applied Psychology, 90: 945955. Aiken, L. S., & West, S. G. 1991. Multiple regression: Testing and interpreting interactions. Newbury Park, CA: Sage. Amabile, T. M. 1988. A model of creativity and innovation in organizations. In B. M. Staw & L. L. Cummings (Eds.), Research in organizational behavior, vol. 10: 123167. Greenwich, CT: JAI. Amabile, T. M. 1996. Creativity in context: Update to the social psychology of creativity. Boulder, CO: Westview. Amabile, T. M., & Conti, R. 1999. Changes in the work environment 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 for measuring leader behaviors. Journal of Organizational Behavior, 21: 249 269. Baer, M., & Oldham, G. R. 2006. The curvilinear relation between experienced creative time pressure and creativity: Moderating effects of openness to experience and support for creativity. Journal of Applied Psychology, 91: 963970. Ball, D. 2003. Security trends in the Asia-Pacific region: An emerging complex arms race. Working paper no. 380, Strategic and Defence Studies Centre, Australian National University. Bolin, M., & Ha renstam, A. 2008. An empirical study of bureaucratic and post-bureaucratic characteristics in 90 workplaces. Economic and Industrial Democracy, 29: 541564. Brislin, R. W. 1980. Translation and content analysis of oral and written material. In H. C. Triandis & J. W. Berry (Eds.), Handbook of cross-cultural psychology, vol. 1: 389 444. Boston: Allyn & Bacon. Bunderson, J. S., & Boumgarden, P. 2010. Structure and learning in self-managed teams: Why bureaucratic

640

Academy of Management Journal

June

teams can be better learners. Organization Science, 21: 609 624. Bunderson, J. S., & Sutcliffe, K. M. 2003. Management team learning orientation and business unit performance. Journal of Applied Psychology, 88: 552560. Burns, T., & Stalker, G. M. 1961. The management of innovation. London: Tavistock. Button, S. B., Mathieu, J. E., & Zajac, D. M. 1996. Goal orientation in organizational research: A conceptual and empirical foundation. Organizational Behavior and Human Decision Processes, 67: 26 48. Caruana, A., Morris, M., & Vella, A. 1998. The effect of centralization and formalization on entrepreneurship in export firms. Journal of Small Business Management, 36: 16 29. Chen, G., & Kanfer, R. 2006. Toward a systems theory of motivated behavior in work teams. In B. Staw & R. M. Kramer (Eds.), Research in organizational behavior, vol. 27: 223267. Greenwich, CT: JAI. Chen, G., Kirkman, B. L., Kanfer, R., Allen, D., & Rosen, B. 2007. A multilevel study of leadership, empowerment, and performance in teams. Journal of Applied Psychology, 92: 331346. DeShon, R. P., & Gillespie, J. Z. 2005. A motivated action theory account of goal orientation. Journal of Applied Psychology, 90: 1096 1127. Dweck, C. S. 1999. Self-theories: Their role in motivation, personality and development. Ann Arbor, MI: Psychology Press. Dyer, N. G., Hanges, P. J., & Hall, R. J. 2005. Applying multilevel confirmatory factor analysis techniques to the study of leadership. Leadership Quarterly, 16: 149 167. Elliot, A. J. 1999. Approach and avoidance motivation and achievement goals. Educational Psychologist, 34: 169 189. Elliot, A. J., & Church, M. A. 1997. A hierarchical model of 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 of correlated method variance in moderated multipleregression analysis. Organizational Behavior and Human Decision Processes, 36: 305323. Farmer, S. M., Tierney, P., & Kung-McIntyre, K. 2003. Employee creativity in Taiwan: An application of role identity theory. Academy of Management Journal, 46: 618 630. Gong, Y., Huang, J. C., & Farh, J. L. 2009. Employee learning orientation, transformational leadership, and employee creativity: The mediating role of employee creative self-efficacy. Academy of Management Journal, 52: 765778. Hage, J., & Aiken, M. 1967. Relationship of centralization

to other structural properties. Administrative Science Quarterly, 12: 7292. Hall, R. D. 1999. Organizations: Structures, processes, and outcomes. London: Prentice-Hall. Harrison, D. A., Price, K. H., & Bell, M. P. 1998. Beyond relational demography: Time and the effects of surface- and deep-level diversity on work group cohesion. Academy of Management Journal, 41: 96 107. Harrison, M. M., Neff, N. L., Schwall, A. R., & Zhao, X. 2006. A meta-analytic investigation of individual creativity and innovation. Paper presented at the 21st Annual Conference of the Society for Industrial and Organizational Psychology, Dallas. Hirst, G., van Knippenberg, D., & Zhou, J. 2009. A crosslevel perspective on employee creativity: Goal orientation, 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 culturesA qualitative and quantitative study across 20 cases. Administrative Science Quarterly, 35: 286 316. James, L. R., Demaree, R. B., & Wolf, G. 1984. Estimating within group interrater reliability with and without response bias. Journal of Applied Psychology, 69: 8 98. Janssen, O., & Van Yperen, N. W. 2004. Employees goal orientations, the quality of leader-member exchange, and the outcomes of job performance and job satisfaction. Academy of Management Journal, 47: 368 384. Kirkman, B. L., & Rosen, B. 1997. A model of work team empowerment. In R. W. Woodman & W. A. Pasmore (Eds.), Research in organizational change and development, vol. 10: 131167. Greenwich, CT: JAI. Klein, K. J., & Kozlowski, S. W. J. 2000. Multilevel theory, research, and methods in organizations. San Francisco: Jossey-Bass. Kristof, A. L. 1996. Person-organization fit: An integrative review of its conceptualization, measurement, and implications. Personnel Psychology, 49: 1 49. Lee, H., & Choi, B. 2003. Knowledge management enablers, 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 difficulties of detecting interactions and moderator effects. Psychological Bulletin, 114: 376 390. Mischel, W. 1977. The interaction of person and situation. In D. Magnusson & N. S. Endler (Eds.), Personality at the crossroads: Current issues in interactional psychology: 333352. Hillsdale, NJ: Erlbaum. Muthe n, B. 1994. Multilevel covariance structure analysis. Sociological Methods & Research, 22: 376 398.

2011

Hirst, van Knippenberg, Chen, and Sacramento

641

Nasser, F., & Wisenbaker, J. A. 2003. A Monte Carlo study investigating the impact of item parcelling on measures of fit in confirmatory factor analysis. Educational and Psychological Measurement, 63: 729 757. Oldham, G. R., & Cummings, A. 1996. Employee creativity: Personal and contextual factors at work. Academy of Management Journal, 39: 607 634. Porath, C. L., & Bateman, T. S. 2006. Self-regulation: From goal orientation to job performance. Journal of Applied Psychology, 91: 185192. Preacher, K. J., Curran, P. J., & Bauer, D. J. 2006. Computational tools for probing interactions in multiple linear regression, multilevel modeling, and latent curve analysis. Journal of Educational and Behavioral Statistics, 31: 437 448. Rafferty, A. E., & Griffin, M. A. 2004. Dimensions of transformational leadership: Conceptual and empirical extensions. Leadership Quarterly, 15: 329 354. Raub, S. 2007. Does bureaucracy kill individual initiative? The impact of structure on organizational citizenship behavior in the hospitality industry. International Journal of Hospitality Management, 27: 179 186. Raudenbush, S. W., & Bryk, A. S. 2002. Hierarchical linear models. Thousand Oaks, CA: Sage. Seijts, G. H., Latham, G. P., Tasa, K., & Latham, B. W. 2004. Goal setting and goal orientation: An integration of two different yet related literatures. Academy of Management Journal, 47: 227239. Shacham, R. 2009. A utility for exploring HLM 2 and 3 way interactions Software available from http:// www.people.ku.edu/preacher/interact/shacham/ index.htm. Shalley, C. E., Zhou, J., & Oldham, G. R. 2004. The effects of personal and contextual characteristics on creativity: Where should we go from here? Journal of Management, 30: 933958. Srivastava, A., Bartol, K. M., & Locke, E. A. 2006. Empowering leadership in management teams: Effects on knowledge sharing, efficacy, and performance. Academy of Management Journal, 49: 1239 1251. Stevens, C. K., & Gist, M. E. 1997. Effects of self-efficacy and goal-orientation training on effective negotiation skills maintenance: What are the mechanisms? Personnel Psychology, 50: 955978. Taggar, S. 2002. Individual creativity and group ability to utilize individual creative resources: A multilevel model. Academy of Management Journal, 45: 315 330. Tett, R. P., & Burnett, D. D. 2003. A personality traitbased interactionist model of job performance. Journal of Applied Psychology, 88: 500 517. Thompson, V. A. 1965. Bureaucracy and innovation. Administrative Science Quarterly, 10: 120.

Tierney, P., & Farmer, S. M. 2002. Creative self-efficacy: Its potential antecedents and relationship to creative performance. Academy of Management Journal, 45: 1137. Tierney, P., Farmer, S. M., & Graen, G. B. 1999. An examination of leadership and employee creativity: The relevance of traits and relationships. Personnel Psychology, 52: 591 620. Van de Ven, A. H., & Ferry, D. L. 1980. Measuring and assessing organizations. New York: Wiley. VandeWalle, D. 1997. Development and validation of a work domain goal orientation instrument. Educational and Psychological Measurement, 57: 995 1015. VandeWalle, D., Brown, S. P., Cron, W. L., & Slocum, J. W. 1999. The influence of goal orientation and self-regulation tactics on sales performance: A longitudinal field test. Journal of Applied Psychology, 84: 249 259. Wagner, J. A. 1994. Participations effects on performance and satisfaction: A reconsideration of the research evidence. Academy of Management Review, 19: 312330. Zhou, J., & Shalley, C. E. 2008. Handbook of organizational creativity. New York: Erlbaum.

Giles Hirst (giles.hirst@monash.edu) is an associate professor and the deputy director of research at Monash University, Department of Management, Faculty of Business & Economics. He received his Ph.D. from the Melbourne Business School. His research interests include the study of employee creativity and innovation. Daan Van Knippenberg (dvanknippenberg@rsm.nl) is a professor of organizational behavior, Rotterdam School of Management, Erasmus University Rotterdam. He received his Ph.D. from Leiden University. His research interests include leadership, diversity, team performance, creativity, and social identity. Chin-Hui Chen (c5286@webmail.customs.gov.tw) is a customs officer at the Kaohsiung Customs Bureau, Taiwan. He is also a part-time lecturer in management at the Fooyin University. He received his Ph.D. in human resource management from the National Sun Yat-Sen University. His research interests include job stress, selfefficacy, psychological well-being, and goal orientation. Claudia A. Sacramento (c.a.sacramento@aston.ac.uk) is an assistant professor of organizational behavior at the Department of Work and Organisational Psychology, Aston Business School. She received her Ph.D. from Aston University. Her research interests include creativity and team effectiveness.

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.