Leadership and innovation in organizations

Leadership and Innovation: How and When do Leaders Influence Innovation in R&D Teams? Leif Denti DEPT OF PSYCHOLOGY and GOTHENBURG RESEARCH INSTITUTE Sweden, 2011 SCHOOL OF TECHNOLOGY AND SOCIETY © Leif Denti Printed in Sweden Department of Psychology and Gothenburg Research Institute University of Gothenburg Gothenburg, 2011 ISSN 1101-718X GU/PSYK/AVH--248--SE DEGREE OF LICENTIATE IN PSYCHOLOGY Abstract Denti, L. (2011). Leadership and Innovation: How and When do Leaders Influence Innovation in R&D Teams? University of Gothenburg. Sweden. This thesis examines how and when leadership is related to individual innovation in organizations, i.e., those factors that mediate and moderate the relationship between leadership and innovation. In study I, 30 years of research into leaders‘ influence on organizational innovation was systematically reviewed to identify these factors. The sample consisted of 27 empirical studies treating leadership as the independent variable and innovation as the dependent variable. A further purpose of study I was to identify new potential mediating and moderating factors. Study II, based on study I, surveyed 166 team members in research and development (R&D) teams, together with their team leaders and department managers, in five Swedish industrial organizations. In this study, leadership was conceptualized using leader–member exchange theory (LMX). The main findings indicated that individual personal initiative predicted individual innovation behavior, while individual intrinsic motivation and LMX did not. Innovation behavior was measured using both quantitative indicators (e.g., new patents, products, scientific publications, and other publications) and leaders‘ ratings. A mediating effect was found in which leader–member exchange was associated with innovation through the personal initiative of team members. Organizational support moderated the relationship between LMX and individual personal initiative, strengthening the relationship when organizational support was high. The two studies concluded that leaders do influence the innovativeness of their teams and employees in various ways. Furthermore, factors residing at the individual, team, and organizational levels of analysis may facilitate or undermine the efforts of leaders to influence innovation outcomes. Keywords: LMX, innovation, mediators, moderators, R&D Leif Denti, Gothenburg Research Institute, University of Gothenburg, Box 603, 405 30 Göteborg, Sweden. Phone: +46 31 786 5608. E-mail: leif.denti@gri.gu.se Populärvetenskaplig svensk sammanfattning Den här uppsatsen inriktas på hur och när ledarskap i avancerade forsknings- och utvecklingsgrupper i svensk industri relaterar till innovationer och stärkt innovationsförmåga för utvecklingsprojekten. Innovationsförmåga, eller förmågan att utveckla nya produkter, tjänster eller patent har lyfts fram som nyckeln till överlevnad för många svenska företag, som är utsatta för en tilltagande internationell konkurrens. Samtidigt har ledarskap under de senaste åren lyfts fram som en kritisk faktor för innovationsskapande. Det saknas emellertid specifik kunskap rörande hur och när ledare kan påverka innovationsförmågan hos innovationsinriktade projektgrupper och dess medlemmar. Med frågan hur menas genom vilka mekanismer ledare stimulerar innovationsförmåga. Dessa mekanismer kallas för medierande faktorer eftersom de bidrar till att förklara sambanden mellan ledarskap och innovation. Till exempel kan en projektledare uppmuntra till öppen diskussion och idégenerering, vilket i sin tur har visat sig öka innovationsförmågan hos medarbetare. Med frågan när menas vilka faktorer som kan påverka ledares inflytande på innovationer i projektgrupper. Dessa faktorer kallas för modererande faktorer eftersom de stärker eller försvagar sambanden mellan ledarskap och innovation. Till exempel har ledare större handlingsutrymme om de befinner sig i ett gynnsamt organisatoriskt klimat där innovation uppmuntras, om ett stort mått av frihet för projektgruppen ges och om resurser finns tillgängliga. Studie I syftade till att systematiskt gå igenom 30 års forskning för att undersöka vilka faktorer som visat sig mediera eller moderera sambanden mellan ledarskap och innovationsförmåga. Underlaget utgjorde 27 empiriska studier över ledarskap och innovationsförmåga. I dessa studier kunde medierande och modererande faktorer identifieras på individuell-, grupp- och organisationsnivå. Ytterligare ett syfte med studie I var att identifiera nya potentiella faktorer som kunde förklara sambanden (mediera) eller påverka sambanden (moderera) mellan ledarskap och innovationsförmåga. Studie II baserades på studie I och testade en modell över medierande faktorer (ledare stimulerar sina gruppmedlemmars motivation samt initiativtagande, vilka i sin tur är positivt associerade med innovationsförmåga) och en modererande faktor (graden av organisatoriskt stöd påverkar ledares möjligheter att stimulera sina gruppmedlemmars motivation samt initiativtagande). Ledarskap operationaliserades med stöd i ‘leader–member exchange‘-teorin (LMX), som ser ledarskap som en arbetsrelation beståendes av sociala utbyten mellan ledaren och dennes enskilda gruppmedlemmar. Denna arbetsrelation kan vara av varierande kvalitet. En lågkvalitativ arbetsrelation utgår från det grundläggande arbetskontraktet, där gruppmedlemmens tid byts mot pengar, medan ledare och medlemmar i arbetsrelationer av högre kvalitet utbyter ömsesidig tillit, respekt, uppskattning och arbetsinsatser gentemot gruppens mål. 166 gruppmedlemmar, deras projektledare och avdelningschefer i 43 avancerade forsknings- och utvecklingsgrupper i svensk industri svarade på ett frågeformulär rörande LMX, organisatoriskt stöd, motivation, personligt initiativtagande och individuellt innovativt beteende. Resultaten från en så kallad path-analys var att personlig initiativförmåga predicerade individuellt innovativt beteende, medan motivation och LMX inte hade direkta samband med detta utfall. Innovativt beteende mättes med både kvantitativa indikatorer, såsom nya patent, produkter, och publikationer, men även med ledarnas subjektiva bedömningar. En medieringseffekt hittades där LMX var indirekt relaterad till individuell innovation via personligt initiativtagande. Organisatoriskt stöd modererade relationen mellan LMX och personligt initiativtagande på så sätt att relationen blev starkare när graden av organisatoriskt stöd var starkt. Från de två studierna dras två huvudsakliga slutsatser. För det första verkar ledare kunna stimulera innovationsförmågan hos sina projektgrupper och dess medlemmar. Till exempel kan de uppmuntra till diskussion och idégenerering, öka medlemmarnas personliga initiativtagande och deras känsla av att kunna ge ett kreativt bidrag. För det andra verkar faktorer som finns på individuell-, team och organisationsnivå underlätta eller minska ledares ansträngningar att stimulera innovationsförmågan hos sina projektgrupper. Vikten av att aktivt stödja innovation lyfts fram som en kritisk faktor, i synnerhet genom att uppmuntra innovation i kommunikation och handling, ge tillräckligt med frihet till utvecklingsgrupper och att tillhandahålla resurser i form av utrustning, information och expertkunskap. Preface This thesis consists of two studies, designated as Study I and Study II. Study I. Denti, L., & Hemlin, S. Leadership and innovation in organizations: A systematic review of factors that mediate or moderate the relationship. Submitted for publication. Study II. Denti, L., & Hemlin, S. Leader-member exchange and individual innovation: Intrinsic motivation and personal initiative as mediators, and the role of organizational support. Manuscript. Acknowledgements First of all, I would like to give my very warmest and special thanks to my supervisor, Professor Sven Hemlin. His expertise and charismatic leadership style was an empowering influence. I also thank Professor Stefan Tengblad, leader of the research project, for valuable input and discussions. I am especially thankful for the high amount of autonomy I experienced under their supervision. Furthermore I would like to thank my co-supervisor Professor Jan Johansson-Hanse with his vast methodological knowledge and sharp analytical sense. I thank VINNOVA (The Swedish Governmental Agency for Innovation Systems) for funding the project, AB Netsurvey for excellent collaboration, the organizations and people involved as participants in this study, my colleagues at the Department of Psychology and Gothenburg Research Institute (GRI) as well as friends and family for bringing me support and joy. Contents Introduction ...................................................................................................................................... 1 The elusive concepts of innovation and creativity ....................................................................... 2 Measuring innovation ................................................................................................................... 3 Factors related to innovation capabilities ......................................................................................... 6 The individual ............................................................................................................................... 8 The team ....................................................................................................................................... 9 The organization ......................................................................................................................... 10 The role of leaders in innovation endeavors .................................................................................. 13 Theories of leadership in relation to innovation ......................................................................... 15 Leader–member exchange theory ............................................................................................... 16 General aim of this thesis ............................................................................................................... 17 How mediating and moderating variables work ......................................................................... 18 Summary of the empirical studies .................................................................................................. 20 Overview and specific aims of the studies ................................................................................. 20 Study I......................................................................................................................................... 20 Study II ....................................................................................................................................... 24 Discussion ...................................................................................................................................... 28 The impact of leaders on team member innovation ................................................................... 28 Moderators of leader impact on team member innovation ......................................................... 29 The central role of individuals in innovation.............................................................................. 31 The influence of Swedish management principles ..................................................................... 31 Further conclusions and implications for organizations ............................................................. 32 Limitations of this thesis............................................................................................................. 34 Recommendations for future research ........................................................................................ 35 References ...................................................................................................................................... 37 Leadership and Innovation: How and When do Leaders Influence Innovation in R&D Teams? Under the pressure of international competition, with increasing demand for more complex and differentiated products and services, developing innovation capability has become a key goal of organizations (Mumford, Scott, Gaddis, & Strange, 2002; Tidd, Bessant, & Pavitt, 2001). Shorter product life cycles, with frequent replacements and improvements in product lines, add to the pressure for innovation as production processes are shortened to meet deadlines (Tidd et al., 2001). In researching numerous organizational settings and management levels, many scholars have tried to identify and explain the factors that advance or hamper innovation and its precursor, creativity. This thesis investigates how leadership is related to innovation, examining the mechanisms by which leaders can stimulate the innovation abilities of their team members. These factors are called mediators, because they explain the relationship between leadership and innovation. The thesis also investigates when leadership is related to innovation, examining the contingency factors that facilitate or hinder leaders‘ efforts. These factors are called moderators, because they strengthen or weaken the relationship between leadership and innovation. 1 The elusive concepts of innovation and creativity The concepts of innovation and creativity are frequently used interchangeably in the literature (Basadur, 2004; Csikszentmihalyi, 1999), and the two concepts can be described as intertwined at best. Simonton (2003) calls our attention to the fact that creativity can be viewed in terms of a tangible product, a personal trait (some individuals being more creative than others), and a cognitive or work process (contrasted with routine or mundane activities). While creativity and innovation together produce original outcomes, creativity precedes innovation in a multistage process (Amabile, Conti, Coon, Lazenby, & Herron, 1996; Scott & Bruce, 1994). Creativity is required at various stages of the process of turning ideas into outcomes, but is only part of the innovation process. Furthermore, innovation, as an implementation-focused process, aims to benefit the organization, though this is not necessarily the goal of creativity (Anderson, De Dreu, & Nijstad, 2004). Creativity is often defined as idea generation, or ideation. According to Amabile et al. (1996, p. 1155), creativity is ―the production of novel and useful ideas.‖ In another definition, creativity expands to include aspects of implementation, being ―the creation of a valuable, useful new product, service, idea, procedure or process‖ (Woodman, Sawyer, & Griffin, 1993, p. 293). Yet some scholars still distinguish the two concepts and argue that creativity is solely ideation, whereas innovation is the process of realizing and applying these ideas (e.g., Anderson et al., 2004; Mumford & Gustafson, 1988; Shalley & Gilson, 2004). This thesis views innovation as ―the implementation of a new or significantly improved product (good or service), or process, a new marketing method, or a new organizational method in business practices, workplace organization or external relations‖ (OECD, 2005, p. 46), and creativity as the generation of novel and useful ideas (Amabile et al., 1996). 2 Measuring innovation If creativity is the precursor to innovation, where does creativity end and innovation begin? This question is especially pressing when the same indicators are used to measure both concepts empirically. For example, both Tierney, Farmer, and Graen (1999) and Scott and Bruce (1994) used the quantitative indicator ―invention disclosures,‖ in measuring creative (defined as the production of novel and useful ideas, i.e., ideation) and innovative (defined as idea implementation, i.e., distinct from ideation) individual behavior, respectively. The above question challenges future research in which creativity or innovation is measured as the outcome. This thesis concerns individual innovation. Below, two approaches to measuring innovation are discussed, which are defined as ―soft‖ and ―hard‖ measures. This thesis defines ―soft‖ measures as those that measure human psychological phenomena (e.g., attitudes and judgments) and ―hard‖ measures as those that measure quantitative indicators of innovation (e.g., patents, products, or scientific publications). For example, the number of new products is a ―hard‖ measure of innovation, whereas someone‘s evaluations of the quality, novelty, and added value of a product are ―soft‖ measures. ―Soft‖ measures used in past research include individual and team self-reports (e.g., Atwater & Carmeli, 2009), supervisor reports on individuals or teams (e.g., Scott & Bruce, 1994), peer reports (e.g., Amabile, Schatzel, Moneta, & Kramer, 2004), and expert or external assessments (e.g., Jung, Wu, & Chow, 2008; Sosik, Avolio, & Kahai, 1997). ―Hard‖ measures used in past research include such indicators as individual invention disclosures and research reports (e.g., Scott & Bruce, 1994), individual implemented innovations (e.g., Rank, Nelson, Allen, & Xu, 2009), new products created in a specific period (e.g., García-Morales, MathíasReche, & Hurtado-Torres, 2008), and new patents. Annual research and development (R&D) 3 expenditure as a percentage of gross revenue (Jung et al., 2008) is also used as a ―hard‖ measure of innovation. We argue that researchers need to use more ―hard‖ measures of innovation for at least three reasons. First, ―hard‖ measures of innovation encompass much of the innovation process and thus can be used as proxies in evaluating innovation performance and technological success (Archibugi & Pianta, 1992, 1996; Fai, 2007; OECD, 1994; Pelz & Andrews, 1966). Second, ―soft‖ measures depend on human judgment and are more easily subject to bias than are ―hard‖ or objective measures. Third, as both ―soft‖ and ―hard‖ measures refer to the same phenomenon, i.e., innovation, ―hard‖ measures can validate ―soft‖ measures and vice versa. On the other hand, it should be borne in mind that ―hard‖ measures are more or less apt proxies for innovation. Table 1 presents examples of ―hard,‖ i.e., quantitative, measures used in innovation research, as well as reports of their convergent validity with other measures. 4 Table 1. Examples of quantitative measures used in innovation research at the individual, team, and organizational levels of analysis. Measure Studies Convergent validity Scott & Bruce (1994) Significant correlation with leader-rated individual innovation behavior (r = .33**) Individual level Invention disclosures Number and effectiveness of Rank et al. (2009) implemented innovations (leader rated) Significant correlation with leader-rated individual innovation behavior (r = .44**) Team level Scientific publications Hemlin, Prpic, & Denti (2009) Other measures were not collected West, Borril, Dawson, Brodbeck, Shapiro, & Haward (2003) Other measures were not collected Number of process innovations Organizational level Number of patents Number of patent citations Jung et al. (2008) Significant correlation with expert ratings of 50 organizations (r = .50a) Makri & Scandura (2010) No correlation with number of patents Ratio of annual R&D spending to annual sales Jung et al. (2008) Ratio of sales of new products to total sales Czarnitzki & Kraft (2004) Other measures were not collected 1. Ratio of sales of new products to total Gumusluoğlu & Ilsev sales, 2. ratio of sales of new products to (2009) R&D expenditures. No correlation with patents or expert ratings, but similar pattern of correlations with independent variables Other measures were not collected Number of 1. product/market Elenkov and Manev (2009) The sub-dimensions of the product/market innovations (e.g., new products and new innovations and organizational markets entered) and 2. organizational innovations loaded on separate factors; innovations (e.g., new planning/control they correlated significantly (r = .42**) systems) adopted by an organization over a two-year period 1. New products, 2. new markets entered, 3. total R&D spending, 4. number of employees in R&D García-Morales, Matias- Significant correlation with CEO Reche, & Hurtado-Torres, subjective ratings of organizational 2008 innovation (r = .88**) a p-value not reported * Correlation is significant at the 0.05 level (two-tailed) ** Correlation is significant at the 0.01 level (two-tailed) 5 Factors related to innovation capabilities The past three decades of innovation research (1980–2011) have arrived at a fairly comprehensible overview of the antecedent factors that facilitate organizationally based innovation, encompassing the domains of the individual, team, and organization (Anderson et al., 2004). However, innovation processes are complex, occurring at various and nested levels of human organizing. In addition, the commercial side of innovation makes for a great demand for more precise knowledge of organizational innovativeness (Kanter, 1996; Paulus & Yang, 2000; Shelley & Perry-Smoth, 2000; Sternberg, 1999; Williams & Young, 1999). This thesis acknowledges that many authors, particularly in the economic disciplines, view innovations as mainly the result of inter-organizational processes (e.g., Carlsson, 1997; Edqvist, 1997). This field of innovation research—i.e., innovation systems—is not examined in this thesis, which solely concerns intra-organizational factors and processes. In a recent review of the state of innovation research, Anderson et al. (2004) identified several key factors facilitating innovation at three levels, i.e., the individual, group, and organizational levels. In addition, several other equally noteworthy analyses have been conducted: Woodman et al.‘s (1993) early conceptual theory of organizational creativity, Ford‘s (1996) multi-level theory of creative action in social domains, and Hemlin, Allwood, and Martin‘s (2008) framework of creative knowledge environments. Other relevant literature reviews and meta-analyses have treated leadership and innovation (Mumford et al., 2002), team and organizational climate (Hunter, Bedell, & Mumford, 2007; Hülsheger, Anderson, & Salgado, 2009), and organizational factors affecting innovation (Damanpour, 1991). An attempt to synthesize and condense this literature, applying it at different organizational levels, is made in Table 2. 6 Table 2. Determinants of innovation at the individual, team, and organizational levels. Characteristic Factor Studies (empirical/meta-analytic) Motivation Openness to experience, conscientiousness (N), autonomy, self-efficacy, proactivity, locus of control, need for achievement Intrinsic motivation, extrinsic motivation (N) Barron & Harrington (1981); George & Zhou (2001); Seibert, Kraimer, & Crant ( 2001) Frese, Teng & Wijnen (1999); Prabhu, Sutton & Sauser (2008). Cognitive ability and style Knowledge and expertise, divergent thinking in combination with convergent thinking Individual Personality Basadur, Graen, & Scandura (1986); Feist & Gorman (1998) Team Structure Heterogeneous composition, cohesion (curvilinear) Keller (2001); West & Anderson (1996) Climate Internal communication, openness, emotional Amabile et al. (1996); Anderson & safety, interpersonal relationships, autonomy, West (1998); Bain, Mann, & participation, idea support, risk-taking, conflict (N) Pirola-Merlo (2001); Ekvall (1996); Isaksen, Lauer, Ekvall, & Britz (2001); Tierney et al. (1999) Leadership/leader Participative, supportive, vision, goal setting, high Mumford et al. (2002); Pearce and traits expertise, intrinsic motivation, problem solving Ensley (2004); Tierney et al. skills (1999); Trevelyan (2001) Task characteristics Challenge, stimulation, pressure (curvilinear) Amabile et al. (1996); Hunter et al. (2007) Organizational Structure Culture Specialization, functional differentiation, internal/external communication, formalization (N), centralization (N) Espoused/enacted support for innovation, experimentation, risk-taking, openness, trust Damanpour (1991); Ekvall (1996) Amabile et al. (1996); Bain et al. (2001); Ekvall (1996); Ekvall & Ryhammar (1999); West & Anderson (1996) Resources Time, money, information, expertise Amabile et al. (1996); Damanpour (1991); Ekvall & Ryhammar (1999) Note: Factors thought to have a negative relationship with innovation are marked ―N.‖ 7 The individual Innovation starts with the creative performance of individuals in their own capacities or together with others. Certain factors pertaining to the individual may be seen as antecedents to the creative process in organizations (Pirola-Merlo & Mann, 2004). In a multifaceted metaanalysis of the personalities of artists and scientists, Feist and Gorman (1998) identified a common pattern for the creative personality, which was characterized by openness, flexibility, self-confidence, high self-efficacy, and a high need for autonomy. The concept of self-efficacy refers to an individual‘s perception of his or her effectiveness in a specific domain. Individuals with a high need for autonomy value individual freedom to pursue their own goals and ideas. Moreover, George and Zhou (2001) were able to establish a relationship between two of the fivefactor personality traits, i.e., openness to experience (positive) and conscientiousness (negative), and creative behavior in the workplace. In addition to these findings, creative behavior at the individual level has been associated with personality traits such as high proactivity (Seibert et al., 2001), high achievement orientation (Barron & Harrington, 1981), and internal locus of control (Woodman et al., 1993). Individuals with an internal locus of control feel that they themselves are in charge of their future, as opposed to individuals with an external locus of control who believe that success or failure is due to factors beyond their control. Some claim that the personality traits that favor creative outcomes are dependent on a key mediating factor—individual intrinsic motivation (Amabile, 1983; Mumford et al., 2002). This concept has been defined as a motivational state generated by the individual in reaction to the inherent challenge of a task (i.e., the work in itself), rather than to extrinsic factors such as rewards (Amabile, 1983, 1998). The motivational state is arguably one of the most important individual factors related to creativity (Amabile, 1983; Woodman et al., 1993). For example, 8 Prabhu et al. (2008) found that intrinsic motivation completely mediated the personality traits of openness to experience and individual sense of self-efficacy, in relation to creative performance. Finally, cognitive ability and style play important roles in creative and innovation performance, since they require knowledge and expertise (Amabile et al., 1996; Hemlin, 2009; Woodman et al., 1993). However, expertise itself does not necessarily lead to creative excellence. Instead, Feist and Gorman (1998) claim that how creative people approach a problem determines the outcome. Highly productive scientists adopt an open and explorative mindset at the beginning of the process, a frame of mind that becomes considerably more incisive, focused, critical, and evaluative toward the end of the process. Divergent thinking, i.e., the ability to combine knowledge elements from diverse sources, is best combined with convergent thinking, i.e., the ability to focus on and select the best solution to a specific problem, to produce creative and innovative outcomes (Woodman et al., 1993). The team Over the years, organizational innovation research has put considerable energy into examining work teams, particularly their composition. Heterogeneous teams, i.e., in which members have diverse skill sets, knowledge, and backgrounds, are arguably more likely to generate innovation outcomes because of the stimulation of divergent thinking in the team (Paulus & Yang, 2000; West & Anderson, 1996). This notion was substantiated by Keller (2001) in a longitudinal study of four manufacturing companies, in which functional diversity indirectly affected the performance of 93 cross-functional research and development teams regarding product technical quality and schedule (time) performance. However, Keller (2001) argued that high team diversity comes with the risk of low cohesion, which may lower innovation capacity. 9 Thus, cohesion may have a curvilinear relationship with innovation performance in teams (Hemlin et al., 2008; Woodman et al., 1993). Furthermore, several factors pertaining to team climate have been linked to innovation performance. Climate can be described as various aspects of the psychological atmosphere in a team and in the surrounding organizational environment (Hemlin et al., 2008). A central climate factor appears to be the level of internal communication and information exchange (Mumford et al., 2002). Studies have demonstrated that highly open communication can influence innovation performance, making team members feel that they can contribute their views on particular ideas without fear of reprisal (Anderson & West, 1998; Bain et al., 2001). In addition, creativity is facilitated in teams in which relationships among team members and with the supervisor are positive and supportive (Isaksen et al., 2001; Tierney et al., 1999), members participate intensely, ideas are supported, risk-taking is accepted, conflict is low (Ekvall, 1996; Hunter et al., 2007; Isaksen et al., 2001), and joy is present (Hemlin, 2009). Last, the design and characteristics of tasks may also affect innovation outcomes. Tasks that are intellectually stimulating and challenging, in contrast to mundane and routine tasks, are thought to spur creative thinking and problem solving (Hunter et al., 2007; Shalley & Gilson, 2004). The organization In a meta-analysis, Damanpour (1991) identified two sets of structural factors determining innovation capability in organizations. First, high degrees of specialization, functional differentiation, and professionalism seem to be conducive to innovation performance, which contributes to the notion that an organization that divides labor based on expertise promotes innovation (Damanpour, 1991). Second, the importance of lively internal and external 10 communication suggests that structures that promote animated, open, and cross-functional communication may help foster innovation. Last, structures that promote centralization and formalization are found to be negatively associated with innovation (Damanpour, 1991). Altogether, the picture that emerges is that creativity and innovation appear to occur more naturally in decentralized, organic, and flexible, rather than mechanistic and rigid, organizational contexts (Kanter, 1996; Mumford et al., 2002; Thompson, 1965). Innovation also somewhat depends on organizational culture, more specifically, on the degree of organizational support. Mann (2005) argued that organizational support can be divided into three forms: 1) organizational encouragement of innovation, i.e., the degree to which researchers feel and perceive various types of support, for example, idea support, trust, emotional safety, and acceptance of risk-taking; 2) access to requisite resources, i.e., primarily time, materials, expertise (know-how), and information; and 3) empowerment, i.e., the extent to which researchers feel encouraged and experience freedom to approach a task, in other words, the degree of autonomy. The presence of these values and resources may lead to actual improvement in innovation performance (Bain et al., 2001; Ekvall & Ryhammar, 1999). In a questionnaire study of management groups in hospitals, West and Anderson (1996) found that organizational support for innovation was the strongest factor predicting overall innovation, i.e., implemented organizational changes. In particular, autonomy, or the freedom granted to pursue unique ideas and insights, has consistently been linked to innovation performance (e.g., Ekvall, 1996; Hunter et al., 2007). Autonomy can empower the group, sending signals of organizational trust that invokes a sense of ownership and control (Amabile, 1998; Mann, 2005; Pirola-Merlo, 2000). Finally, from an organizational perspective, innovation is often resource intensive. Several researchers have proposed that allotting sufficient resources may be a determining factor for innovativeness (e.g., Damanpour, 1991; Mumford et al., 2002; Woodman et al., 1993). 11 In summary, organizations wanting to maximize their innovation capabilities should first assess their structural prerequisites. Overly formalized and bureaucratic organizational structures seem to impede innovation; in contrast, organizational structures in which decision making and influence regarding organizational processes are decentralized and in which project teams are granted considerable autonomy have been found to facilitate innovation (Damanpour, 1991; Jung et al., 2008; Thompson, 1965). An innovation policy should be implemented that encourages, expects, and rewards creative ideas (Mumford & Gustafsson, 1988), promotes open discussion (i.e., in which ideas and criticism can be expressed without repercussions), and accepts failure (Mann, 2005; Pirola-Merlo, 2000). Second, project teams should be composed with heterogeneity of competences in mind (Reiter-Palmon & Illies, 2004), and consist of members with creative and proactive personalities (Feist & Gorman, 1998; Seibert et al., 2001). Teams and their individual members should furthermore be assigned tasks that are challenging and stimulating, and that increase intrinsic motivation (Amabile, 1998; Oldham & Cummings, 1996; Shalley & Gilson, 2004). However, a fundamental question must be asked: Where are the ultimate responsibilities found for implementing these principles and creating environments conducive to innovation? The answer lies with the conductors of the symphony called organizational innovation—the leaders. 12 The role of leaders in innovation endeavors Given the vast body of creativity and innovation research, little attention has been paid to the relationship between innovation and the leadership domain (Byrne, Mumford, Barrett, & Vessey, 2009), especially in research and development environments (Elkins & Keller, 2003). In complex systems such as organizations, leader impact on innovation outcomes has been viewed as only one of several influences (Kaiser, Hogan, & Craig, 2008). For example, progress in innovation endeavors is often non-linear, leaving significant room for unpredictable dynamics (Marion & Uhl-Bien, 2001), including the influence of external forces and chance (Kaiser et al., 2008). From this perspective, innovation has been viewed as a complex process that simply cannot be adequately and systematically managed (Tidd et al., 2001). Another reason may be the ―romantic conception of the creative act‖ (Mumford et al., 2002, p. 706), according to which creative ideas are conceived by individuals to which the supervisor constitutes an obstacle rather than a facilitator (Jung, 2001). However, conclusions from research into intra-organizational innovation point in the opposite direction: Leaders are increasingly being recognized as essential in facilitating innovation because they can create the conditions and circumstances needed for creativity and innovation to flourish (Kaiser et al., 2008; Mumford et al., 2002; Shalley & Gilson, 2004). As a general framework, this thesis conceives of leaders as integral to organizational innovation performance in two ways, which can be described as the dual process of managing innovation. First, leaders are central in creating the context and opportunities for teams and employees that favor creativity and ultimately innovation (Shalley & Gilson, 2004), for example, by creating and supporting a positive team climate (Anderson & West, 1998), by facilitating problem-solving and team reflection (Puccio, Murdock, & Mance, 2007; Somech, 2006; Tierney 13 et al., 1999), and by creating team heterogeneity (Keller, 2001). Moreover, leaders can increase individual intrinsic motivation (Avolio, Bass, & Jung, 1999) and establish and maintain highquality work relationships with team members (Scott & Bruce, 1994). In an organizational context, the leadership role may be seen as enabling innovation as a bottom–up process in which the leader acts as a facilitator creating the conditions for team members to use their capacities in producing creative innovation outcomes. Second, leaders embody the organization‘s desires to become innovative, constituting one of the primary channels by which these desires can be realized, for example, by managing and allocating resources in the form of time, facilities, money, and knowledge (Drazin, Glynn, & Kazanjian, 1999), setting and managing individual and team goals (Shalley & Gilson, 2004) and expectations for innovation performance (Yuan & Woodman, 2010), managing rewards (Mumford & Gustafson, 1988), and granting autonomy to individuals and teams (Hemlin, 2006; Hülsheger et al., 2009). Accordingly, this could be viewed as a top–down process, in which the leader manages the strategic innovation goals of the organization. The leader could be seen as a focal point or hub in orchestrating the dual processes of a) creative teams or individuals wanting to turn their creative efforts into innovation outcomes (leader as facilitator) and b) the goals and efforts of the organization to be innovative (leader as manager). 14 Theories of leadership in relation to innovation Leadership is a multifaceted concept with no single definition. The most influential definitions state that leadership is a process by which one individual exerts influence over a group (Strannegård & Jönsson, 2009). Leadership in organizations is typically studied at the individual level, for example, in research or project teams consisting of a formal leader and a number of team members. This thesis focuses on the formal leaders of such teams. The most influential leadership theory is the transformational/transactional theory of leadership (e.g., Bass, 1985), which has been related to innovation performance both when studying individuals and teams (e.g., Eisenbeiss, van Knippenberg, & Boerner, 2008; Rank et al., 2009) and when whole organizations are the unit of measurement (e.g., Aragón-Correra, GarcíaMorales, & Cordón-Pozo, 2007; Gumusluoğlu & Ilsev, 2009; Jansen, Vera, & Crossan, 2009). Transformational leaders exert influence by ―broadening and elevating followers‘ goals and providing them with confidence to perform beyond the expectations specified in the implicit or explicit exchange agreement‖ (Dvir, Eden, Avolio, & Shamir, 2002, p. 735). In contrast, transactional leaders exert influence by means of the contractual exchange of rewards and corrective actions (Avolio et al., 1999). Other studies have operationalized leadership as various leader behaviors that are positively related to individual and team innovation, for example, behaviors that aim at clarifying problem construction and improving self-efficacy (Redmond, Mumford, & Teach, 1993), stimulating open discussion and debate (Somech, 2006), and providing support and encouragement (Krause, 2004; Oldham & Cummings, 1996). Leadership comprises three elements: the leader, the group members, and the leader– member work relationship (Graen & Uhl-Bien, 1995). While much past research into the leader‘s role in innovation has focused on the style and behaviors of the leader, this thesis views 15 leadership from the relationship perspective. This thesis conceptualizes the leader-member work relationship with leader-member exchange (LMX) theory (Graen & Cashman, 1975; Graen & Uhl-Bien, 1995), which is viewed as the primary conduit through which leader influence is exerted. Leader–member exchange theory Leader–member exchange theory, theoretically rooted in role theory and social exchange theory, is distinguished from other leadership theory by its focus on the unique work relationship between supervisor and employee or, in this thesis, between team leader and member (Gerstner & Day, 1997). LMX theory views leadership as a tacit agreement regarding what is expected from each participant in the leader–member dyad. Low-quality LMX relationships are based primarily on the employment contract, the interaction being formal and impersonal, while in high-quality LMX relationships, team members and leaders go beyond the formal terms and conditions of the work contract, to a state in which interactions are based on mutual trust, respect, liking, and influence (Greguras & Ford, 2006). The quality of leader–member exchange has been demonstrated to be predictive of outcomes at the individual, team, and organizational levels of analysis (Gerstner & Day, 1997), outcomes such as work performance (Burton, Sablynski, & Sekiguchi, 2008; Wang, Law, Hackett, Wang, & Chen, 2005), organizational commitment behavior (Burton et al., 2008; Sherony & Green, 2002), and employee job satisfaction and wellbeing (Hooper & Martin, 2008). It has been suggested that the quality of the LMX relationship between leader and team member relates to innovation and creativity. Leaders in high-quality LMX relationships may provide more opportunities for team member autonomy (Liden & Maslyn, 1998), increasing their discretion to implement creative ideas (Hemlin et al., 2008). Furthermore, Mumford and 16 Gustafson (1988) proposed that high-quality LMX relationships influence team member creativity because of the heightened sense of advocacy and trust. Innovation behaviors may be further stimulated by higher expectations of performance and by the fact that team members gain recognition in the organization (Yuan & Woodman, 2010). However, empirical studies of LMX and of innovation and creativity have been scarce, and only a few studies have found positive relationships between LMX and team member innovation (e.g., Basu & Green, 1997; Scott & Bruce, 1994; Yuan & Woodman, 2010). This thesis contributes to this field of research. General aim of this thesis As has been discussed here, antecedents to organizational innovation have been identified at the individual, team, and organizational levels. As the research field of creativity and innovation in R&D can be described as still fragmented, and science has not yet settled on a taxonomy of factors affecting these processes and how these factors interact (Hemlin et al., 2008), several authors are advocating more inquiry (e.g., Anderson et al., 2004; Hemlin et al., 2008; Hunter et al., 2007). Regarding leadership, our knowledge is particularly lacking when it comes to contextual factors that affect leaders‘ abilities to influence processes of innovation in organizations. This thesis answers the call of many scholars (e.g., Avolio, 2007; Byrne et al., 2009; Graen & Ulh-Bien, 1995; Hackman & Wageman, 2007; Mumford et al., 2002; Shalley & Gilson, 2004) to investigate the factors moderating and mediating between leadership and innovation. 17 How mediating and moderating variables work It is important to differentiate between mediation and moderation. Both mediating and moderating variables are ―third variables‖ that explain some aspect of the relationship between an independent variable, or predictor, and a dependent variable, or criterion. In this thesis, the predictor variable is leadership and the criterion variable is innovation. Mediators. A mediating variable is defined by Baron and Kenny (1986, p. 1176) as a variable that ―accounts for the relation between the predictor and the criterion.‖ For example, we may take the predictor ―smoking‖ and the criterion ―premature death.‖ If we assume that this relationship is positive and true, i.e., if one smokes more, then one has an increased risk of dying prematurely, a mediator variable such as ―lung cancer‖ would explain this relationship. If one smokes more, the chance of developing lung cancer increases, which in turn increases the risk of premature death. Thus, a mediating variable explains the mechanisms involved in a relationship between two other variables (see Figure 1). Predictor Criterion variable variable Mediator variable Figure 1. Path diagram of a single-mediator model. The mediator variable explains an aspect of the mechanism operating between the predictor and the criterion. Moderation. A moderating variable is defined by Baron and Kenny (1986, p. 1174) as ―a qualitative (e.g., sex, race, class) or quantitative (e.g., level of reward) variable that affects the 18 direction and/or strength of the relation between an independent or predictor variable and a dependent or criterion variable.‖ Taking the example of smoking (predictor) and premature death (criterion), if one engages in regular exercise, this would likely affect the smoking–premature death relationship, in that more exercise (with its health benefits) would make this relationship weaker for people who exercise. Hence, exercise is a moderating variable. A moderating variable therefore explains the contingencies or circumstances when a relationship between two other variables is strong or weak (see Figure 2). Moderator variable Predictor Criterion variable variable Figure 2. Path diagram of a single-moderator model. The presence of the moderator variable strengthens or weakens the relationship between the predictor and the criterion. In this thesis, the mediating factors explain how leaders can facilitate or suppress innovation performance., i.e., they explain the relationship between leadership and innovation. The moderating factors explain when leaders can influence innovation, i.e., when they facilitate or suppress leaders‘ efforts to promote innovation. 19 Summary of the empirical studies Overview and specific aims of the studies The first constituent study (study I) of this thesis is a literature review of the field of leadership and innovation studies. This study specifically identified and analyzed the factors known to mediate or moderate the relationship between leadership and innovation performance, and suggested new factors. The second study (study II) empirically tested a model in which leader–member exchange was hypothesized to be positively related to individual innovation. Specifically, this study integrated and tested several streams of research into factors mediating and moderating leaders‘ influence on individual innovation. Leadership, operationalized as the leader–member exchange (LMX), was hypothesized to be mediated by team members‘ intrinsic motivation and personal initiative to advance individual innovation outcomes. Organizational support was hypothesized to moderate the relationship between LMX and team members‘ intrinsic motivation and personal initiative. Study I Study I was a systematic review of the last 30 years of research into factors moderating or mediating the relationship between team leadership and team or individual innovation. Materials and methods. A broad search for empirical articles using the keywords leadership, innovation, and creativity was conducted in online databases. A total of 99 peerreviewed articles were found that satisfied the following criteria: the article had to i) be based on an empirical study and ii) treat leadership as an independent variable and innovation as a dependent variable. We then introduced several quality criteria (e.g., each article had to be 20 published in a journal with a journal impact factor1 >1.0) that reduced the sample size. The articles were coded according to the following categories: i) study sample, ii) level of analysis, iii) type of dependent variable2 (i.e., creativity or innovation) and how it was measured (i.e., using soft or hard measures), iv) independent variables (i.e., how leadership was measured), v) mediating and moderating variables, and vi) results. The final sample consisted of 27 articles. Results of the analysis were presented at the individual, team, and organizational levels. Results. At the individual level, leaders may stimulate their employees‘ creative selfefficacy, i.e., their perception of their propensity to be creative, leading to innovation behaviors (Gong, Huang, & Fahr, 2009; Redmond et al., 1993). Creative self-efficacy therefore mediates the relationship between leadership and individual innovation. It has been suggested that high levels of self-efficacy may contribute to increased motivation (Ford, 1996), greater eagerness to pursue one‘s unique ideas, and more effective use of cognitive resources (Thomas & Velthouse, 1990). Moreover, two factors were found to moderate this relationship. The first such factor is organizational-based self-esteem (OBSE), defined as an individual‘s self-perceived value as an organizational member (Pierce, Gardner, Cummings, & Dunham, 1989). Employees with low OBSE doubt whether their ideas or efforts benefit the organization; OBSE moderates the relationship between leadership and individual innovativeness, since the relationship is more important to employees with low OBSE (Rank et al., 2009). Second, individual self-presentation orientation may also moderate the relationship between leadership and individual innovation. This concept refers to the extent to which an individual engages in behaviors, such as impression management, in order to meet the expectations of the social context in the organization 1 The journal impact factor (IF) of a journal reflects the average number of times each paper in the journal has been cited over the preceding two years. Thus, when a journal has an IF of 1 or more, each article in the journal has, on average, been cited one or more times over the preceding two years (Garfield, 2006). 2 The literature search and coding procedure also served as the basis for another article on factors mediating and moderating the relationship between leadership and creativity (Denti & Hemlin, 2011). The present article includes only those studies that were coded as having innovation as a dependent variable. 21 (Gangestad & Snyder, 2000). High self-monitors, i.e., those with a high self-presentation orientation, tend to control and alter their behavior to present an image congruent with others‘ expectations, while the opposite is the case in individuals low in self-monitoring (Day, Schleicher, Unckless, & Hiller, 2002). Rank et al. (2009) hypothesized and found that selfpresentation moderates the relationship between transformational leadership and individual innovativeness, the relationship being stronger for employees with a low self-presentation orientation. The authors reasoned that, since low self-monitors perform best when involved in an organization in congruence with their beliefs and when they can do things their own way, they will perform at their best under transformational leaders, who consider their employees‘ individual strengths and needs. At the team level, leaders who introduce norms that promote debate, open communication, and divergent thinking may stimulate team innovativeness (Somech, 2006). The communicative process by which team members collectively reflect on the team‘s goals, strategies, and processes was conceptualized by Somech (2006) as ―team reflection.‖ Somech (2006) found that leaders affected team reflection, which in turn affected team innovation, meaning that team reflection acted as a mediator. Furthermore, team heterogeneity was found to moderate leaders‘ efforts when innovation was the goal. When team heterogeneity is low, a more direct style of leadership may be needed to stimulate discussion and divergent thinking, while a more participative leadership style was beneficial when team heterogeneity was high (Somech, 2006). Moreover, task characteristics may act as a moderating variable. Oldham and Cummings (1996) found that when task complexity was high and supervision was non-controlling and supportive, more patents were produced than in other situations. At the organizational level, the relationship between leadership and innovation was found to be strongest in organizations with supportive cultures that encourage innovation in their 22 communications and, most importantly, provide sufficient resources and autonomy to teams that have innovation goals (Howell & Avolio, 1993; Jung et al., 2008). Moreover, organizations that are structurally decentralized and in which formalization is low may provide a more favorable environment for leader innovation because of increased autonomy and inter-functional and interdepartmental collaboration (Jung et al., 2008; Miller, Dröge, & Toulouse, 1988). In such environments, it is more likely that employees can depart from established practices without negative consequences (Dougherty & Hardy, 1996). Furthermore, the study identified two factors for which the findings were too ambiguous to be able to infer whether they act as mediators or moderators in the relationship between leadership and innovation, namely, psychological empowerment at the individual level and team climate at the team level. New moderating and mediating factors between leadership and innovation. Several new moderators and mediators were suggested. At the individual level, an individual‘s number of external work contacts and degree of personal initiative were suggested to mediate the relationship between leadership and innovation. At the team level, team developmental stage was suggested as a moderating factor, and leaders‘ use of various creativity-stimulating techniques as a mediating factor. Finally, the organization‘s reward system was proposed as a moderator at the organizational level. Conclusions. Study I concludes that leaders do have an impact on the innovation abilities of their teams and team members. Organizations that want to lay a foundation for innovation should implement an innovation policy that rewards creative contributions and encourages risktaking and innovation, creating an environment that stimulates individual willingness to undertake creative endeavors (Hemlin et al., 2008; Mumford & Gustafson, 1988). Teams engaged in innovation work should be granted sufficient autonomy for creative problem solving, 23 time limits should be set in the implementation phases (but not too strictly), and teams should be composed with member heterogeneity in mind. Finally, leaders should promote team norms that emphasize open discussion, emotional safety, respect, joy, and creative tension. Study II Study II modeled the relationship between leadership, conceptualized as the leader– member exchange relationship (LMX), and team member innovation, including several factors that moderate and mediate this relationship. Hypotheses. We hypothesized that LMX would be positively related to innovation, and that team members‘ intrinsic motivation and personal initiative would each mediate this relationship. Amabile (1983) has made a strong case that intrinsic motivation is an antecedent to employee creativity, and intrinsic motivation has been found to mediate the relationship between leadership and creativity (e.g., Shin & Zhou, 2003), though it has not been tested in conjunction with individual innovation. Personal initiative, defined as ―a behavior syndrome resulting in an individual‘s taking an active and self-starting approach to work and going beyond what is formally required in a given job‖ (Frese, Fay, Hilburger, Leng, & Tag, 1997, p. 140), may predict innovation (Rank, Pase, & Frese, 2004). Personal initiative may mediate the relationship between leadership and innovation because leaders are assumed to influence their team members‘ propensity to take the initiative (Frohman, 1999). Last, it was hypothesized that organizational support (OSIQ) positively predicts team members‘ intrinsic motivation and personal initiative, and that OSIQ interacts with LMX, strengthening the relationships of LMX with intrinsic motivation and personal initiative, respectively, when organizational support was high. Materials and method. We surveyed 43 R&D-intensive teams in five Swedish industrial corporations. The sample consisted of 166 team members (chiefly scientists and engineers), 43 24 leaders, and 10 department managers. In each team, five team members completed a survey about their work relationships with their team leader, their intrinsic motivation and degree of personal initiative, and the perceived organizational support. Team leaders completed a survey about their work relationship with each team member. We used two strategies to measure innovation. First, we created a quantitative measure by averaging the total number of i) patents, ii) new products, iii) scientific publications, and iv) other publications (e.g., product descriptions and in-house reports) to which a team member had contributed since joining a team under its current leader. We asked team members to report their individual scores on the indicators. Each team leader and department manager reported the total scores on these indicators for the team. Next, a rating scale was used, developed by Scott and Bruce (1994), in which team leaders were prompted to rate their team members and department managers to rate their teams. The quantitative and subjective measures indicated significant medium to high correlations through all three levels of analysis, strengthening the convergent validity of the measures. In further analyses, we used the individual-level quantitative measure provided by members and the subjective measure from the team leaders because our level of inquiry was the individual level. The results were analyzed using path analysis, a method by which the researcher can specify a single model enabling the simultaneous analysis of an entire set of hypotheses. Path analysis also lets us closely examine mediator and moderator mechanisms when analyzing the relationship between independent and dependent variables. Results. The results are presented in Figure 3. When using both measures of innovation, personal initiative was positively related to innovation, while, contrary to our hypotheses, LMX and intrinsic motivation were not directly related to team member innovation. LMX was found to positively predict intrinsic motivation and personal initiative. 25 In assessing the hypothesized mediation effects, personal initiative was found to mediate the relationship between LMX and team member innovation, but the mediator effect was not found for intrinsic motivation, contrary to the hypothesis. LMX was therefore indirectly related to team member innovation, mediated by personal initiative. In the analysis of moderation, a moderator effect was found in which the relationship between LMX and personal initiative was stronger when organizational support was higher. Contrary to what was hypothesized, this moderation effect was not found for organizational support and intrinsic motivation, i.e., the relationship between LMX and intrinsic motivation was not moderated by organizational support. Finally, when using only the self-reported quantitative innovation measure, the control variables gender and group tenure (e.g., length of time as team member) were positively related to team member innovation. Longer-serving group members and men reported more innovations. 26 .17* (.17*) Organizational support (OSIQ) .09 (.08) .10 (.10) .10 (.06) OSIQ x LMXb .20** (.20*) Intrinsic motivation Employee innovation .11 (.11) Personal initiative .22** (.21**) LMX .14 (.09) .10 (-.01) .22** (.22*) .27** (.27**) .01 (.03) Gender Education Group tenure Age .18* (-.05) .09 (.14) .15* (.11) -.07 (.11) Figure 3.a Results for the hypothesized paths between leadership and innovation outcomes mediated by intrinsic motivation and initiative and moderated by organizational support. a Two sets of parameter estimates are presented. The first set (model 1) is when the dependent variable, employee innovation, is the quantitative measure, the second set (model 2) is marked in parentheses, and this is when the dependent variable is the leaders‘ innovativeness ratings. Standardized beta coefficients are given for the structural paths. Covariances are indicated by curves. R2 is presented for the endogenous variables. b This is the interaction term, which was calculated by standardizing and multiplying the OSIQ and LMX scales. * Correlation is significant at the 0.05 level (two-tailed) ** Correlation is significant at the 0.01 level (two-tailed) 27 Conclusions. The main conclusion of study II was that the leading roles in creating innovation outcomes in Swedish industrial R&D are played by the individual engineers and scientists who constitute the work teams, but that leaders may also play a crucial role. The work relationship is suggested to be a hygienic factor, i.e., as long as the relationship is good enough, team members will find themselves satisfactorily supported in their innovation endeavors. It was further concluded that organizations should support innovation by espousing and enacting proinnovation policies, and by providing their teams with sufficient autonomy and resources. This support will make it easier for leaders to create opportunities for their team members to take initiatives and come up with their own ideas and solutions to problems. Discussion The main purpose of this thesis was to examine the influence of leaders of research and development teams on the innovation performance of such teams. The thesis focused on the how, i.e., the mechanisms through which leaders may stimulate the innovation abilities of their team members, and on the when, i.e., the contingency factors that may facilitate or hinder leaders‘ efforts. The impact of leaders on team member innovation First, combining the results of studies I and II, one can conclude that leaders affect the innovativeness of their teams. The literature review in study I, based on current knowledge of team leaders‘ influence on innovation processes, demonstrated that leaders can exert such influence by stimulating discussion and reflection in teams, counteracting narrow and conformist thinking and facilitating innovative ideas (Somech, 2006). Moreover, leaders can stimulate their team members‘ perception of themselves as able to be creative, i.e., their creative self-efficacy, 28 which has been demonstrated to lead to innovation outcomes (Redmond et al., 1993). Research into creative outcomes finds that when leaders express expectations that their teams and employees can be creative, this positively affects their creative performance (e.g., Tierney & Farmer, 2004; Unsworth, Wall, & Carter, 2005). The idea that people tend to behave in accordance with others‘ expectations is called the Pygmalion effect (Eden, 1984), which may be an important tool for leaders seeking to stimulate their employees‘ creative and innovative abilities. Last, study II, based on empirical data on Swedish industry teams, demonstrated that the relationship between leader–member exchange and innovation was mediated by the personal initiative of team members, suggesting that when leaders and team members are involved in highquality work relationships, leaders can provide more opportunities and discretion for team members to be innovative. Moderators of leader impact on team member innovation The second main conclusion of this thesis is that certain factors residing at the individual, team, and organizational levels may either facilitate or hinder leaders‘ efforts to promote innovation in their teams. Study I concluded that the relationship was strongest in organizations that are decentralized and de-formalized, because such organizations grant teams and employees more freedom and discretion to approach their work creatively (Jung et al., 2008). Furthermore, leaders work best under conditions that explicitly support innovation, i.e., where the organization encourages open discussion and risk-taking, grants sufficient autonomy to teams and employees, and provides them with sufficient resources, such as facilities and materials, information, and expertise (e.g., Hunter et al., 2007; Mann, 2005). 29 The role of organizational support. In study II, organizational support moderated the relationship between LMX and team members‘ personal initiative by strengthening this relationship when organizational support was high. LMX theory states that leaders and team members engage in an ongoing process of mutual exchange in the interest of a higher-quality work relationship (Liden & Maslyn, 1998). When organizational support is high, leaders can reward and encourage excellence because they have more options for meeting team member demands. In contrast, in organizational contexts in which support for innovation is weak and innovation is not encouraged, leaders may have difficulties motivating and inspiring innovation. Moreover, leaders in less supportive organizational contexts may find themselves in positions in which they have little discretion when supporting their team members in their work. Thus, as study II has demonstrated, the degree of active innovation support provided by an organization affects innovation in the organization‘s teams, because leaders find it appropriate and easier to encourage team members‘ initiatives. Last, study I identified two moderating factors at the individual level. Leaders seem to have limited influence on individuals with a high propensity for self-monitoring, i.e., who control the images of themselves to better fit in with the organization and the social climate (Rank et al., 2009). As these individuals are more concerned with fitting in than with being forces for innovation, they may cease making attempts if they face resistance to their ideas and suggestions. Moreover, individuals who already feel included in an organization may be less inclined to respond to stimulation from leaders (Rank et al., 2009). These individuals already have a high sense of self-esteem and feel that they are valued as organizational members (Pierce et al., 1989). Findings like these pertaining to moderating variables at the individual level remind us that, while leadership may be important, it is primarily the individual members of work teams who do the actual innovation work. 30 The central role of individuals in innovation Study II found that the direct relationship between leader–member exchange and team member innovation was not significant. While management trends and literature have recently emphasized and somewhat romanticized the role of leaders (Kaiser et al., 2008), this result reminds us that innovation work processes in industrial development teams are carried out by skilled engineers and that leaders may have a less direct role in steering innovation outcomes. The results of study II pertaining to the indirect relationship between leader–member exchange and team member innovation may be interpreted in light of factors that reside at two levels of analysis, the individual and the organizational. First, the sample consisted chiefly of scientists and engineers involved in high-technology innovation. Feist and Gorman (1998) concluded in a meta-analysis that scientists are more driven and have higher needs for achievement and independence. Furthermore, the scientists and engineers in the sample are likely to be highly and intrinsically motivated because of their training and the inherent complexity and challenge of their tasks (Amabile, 1983). Taken together, leaders‘ influence on innovation work may be a hygienic factor for these highly skilled individuals. As long as the leader–member work relationship is not counterproductive, a good enough relationship will ensure that team members have sufficient work support. If this holds true, it also means that a higher LMX relationship will not necessarily lead to higher innovation performance. This is in agreement with Tierney et al. (1999), who found a difference between less and more innovative employees, the latter being little influenced by leaders. The influence of Swedish management principles The second layer of interpretation pertains to the national context of the study. The results of study II may reflect the fact that the sampled team leaders and members were from Sweden, 31 whose organizational management principles differ from those of other national contexts (e.g., Byrkjeflot, 2003; Gerstner & Day, 1994; Holmberg & Åkerblom, 2006). For example, Swedish organizations have been subjected to strong decentralization in recent decades, considerable influence on organizational processes having been transferred to middle managers and the employees themselves (Tengblad, 2003). Democratization influences have emphasized values such as individual and team egalitarianism and autonomy, and leadership based on high employee participation and shared decision making (Holmberg & Åkerblom, 2006). The findings of this thesis, which focuses on individual characteristics and argues that leaders play an important but indirect role, may partly reflect these Swedish managerial principles. Further conclusions and implications for organizations The paradox of autonomy and control. Innovation in organizations is an endeavor fraught with risks, as ideas and projects may misfire and advances seldom follow as intended (Mumford et al., 2002). To eliminate and control these risks, organizations may be tempted to increase control over their innovation projects, for example, by closely monitoring project process parameters, setting stringent time constraints, ending projects prematurely, or employing project ―gates,‖ i.e., specific periods within which a predetermined amount of progress must be made. The need to manage innovation creates a paradox in which increased control constrains the autonomy of teams and employees, which encourages tested-and-true ways of solving problems and stifles the emergence of new ideas that may spawn new innovation projects. The degree of team and team member autonomy has been found to be among the strongest predictors of innovation performance (e.g., Ekvall, 1996; Mann, 2005; West et al., 2003). Still, innovation processes cannot be left completely unmanaged, as some limits and specifications are needed for 32 the successful advance of innovation projects (Tidd et al., 2001). It seems vital that a balance be struck between outright control and laissez faire styles when managing innovation. An innovation policy. One action an organization can take is to implement an innovation policy that explicitly supports and encourages innovation. Risk-taking is inherent to innovation, and organizations that espouse and, more importantly, implement values such as experimentation create a hotbed for innovation (Mann, 2005; Mumford et al., 2002). An organizational culture that encourages innovation and individual creativity signals trust; this may influence individuals‘ willingness to undertake creative endeavors, because they feel that such actions are supported and not undermined (Mumford & Gustafson, 1988; Shalley & Gilson, 2004). The increased autonomy and trust may resonate with individuals who are highly motivated and skilled, making it easier for them to contribute to their organizations‘ innovation goals. Implications for recruiters. Furthermore, the results of this thesis have practical consequences for recruitment policies and team composition in innovation work. Teams should be composed with heterogeneity of competencies in mind and consist of highly motivated individuals rich in initiative. These particular requirements pose a recruitment challenge for human resource management. However, teams like these require competent leadership that is participative and non-controlling, providing individual and group stimulation only when needed, which calls for leaders who are wise, adaptable, and sensitive to the phases of the innovation process (Somech, 2006). 33 Limitations of this thesis The inferences made in this thesis should be viewed in light of its limitations. The primary weakness concerns an effect referred to as ―the graveyard effect,‖ which may have influenced studies I and II and the thesis as a whole. In study I, several quality criteria used in selecting articles (e.g., each article had to be published in a journal with an impact factor above 1 and be indexed in the ISI Web of Science) effectively cut the number of articles in the first selection by two-thirds. The articles excluded from the sample can be considered in the ―graveyard,‖ and we do not know what the implications and inferences would be if all studies were allowed to constitute the sample in study I. This indirectly affected study II, since conclusions from study I were tested in study II. However, it can be argued that imposing fairly strict quality criteria makes the conclusions of this thesis stronger and better substantiated, since articles published in journals with higher impact factors and indexed in the ISI Web of Science are subject to a more rigorous peer-review process, screening out lower-quality research. The second weakness was that a cross-sectional design was used in study II, which limits the causal inferences that can be made. As discussed in study II, some hypothesized relationships may be reciprocal. For example, already highly innovative organizations and project teams may demand certain behaviors from their leaders, such as granting autonomy and time for idea generation and problem solving. Another relationship that may be reciprocal is the relationship between leader–member exchange and intrinsic motivation, because highly motivated individuals may positively influence the work relationship. On the other hand, the relationships hypothesized in study II have a theoretical and empirical basis, which strengthens the causal plausibility of the model. Taking the abovementioned relationship as an example, Deci and Ryan (1987) review 34 causal evidence linking leaders‘ behaviors to employees‘ intrinsic motivation. Still, the path model proposed in study II should be considered only an ―as if‖ model of causality (Kline, 2005). Third, this thesis does not discriminate between radical and incremental innovation. The former concept refers to the creation of new and valuable products, while the latter concerns the improvement and refinement of existing products (Tidd et al., 2001). In study II, the innovation measures were composites of radical and incremental innovation. The processes involved in each form of innovation likely differ, adding a layer of complexity to the conclusions of this thesis. Recommendations for future research This thesis has introduced to the field of innovation research a promising construct, personal initiative, relating to team member innovation performance. However, further research is needed, especially longitudinal research, before we can infer that the relationships are causal. Further research is also needed into the individual construct intrinsic motivation, which was not found to be related to team member innovation, despite a sound theoretical basis and positive relationships with the adjacent concept of individual creativity in other studies (e.g., Shin & Zhou, 2003). However, in innovation work, which concerns activities that go beyond idea generation, such as idea development, idea championing, securing resources, and taking steps to implement ideas concretely (Tidd et al., 2001), personal initiative and the ability to be proactive may be more pertinent than intrinsic motivation. Furthermore, the leader–member work relationship was not directly related to team member innovation in study II, in contrast to the findings of Scott and Bruce (1994) and Yuan and Woodman (2010). Instead, LMX was indirectly related to team member innovation mediated through the personal initiative of team members. This result calls for further research in two ways. First, further research is needed into the role of leaders in innovation ventures in high- 35 technology R&D contexts, in which scientists and engineers are usually highly autonomous, driven (Feist & Gorman, 1998), and intrinsically motivated (Amabile, 1983). This thesis proposes that leaders‘ influence may be a hygienic factor in these contexts. Second, further research is also needed into the influence of cultural factors, as the results in study II may be influenced by cultural artifacts pertaining to Swedish management strategies. Several scholars have called for more cross-cultural research into leadership in innovation ventures (e.g., Anderson et al., 2004; Shalley, Zhou, & Oldham, 2004; Yuan & Woodman, 2010). In the end, as we are only beginning to unravel leaders‘ complex roles in organizational innovation. There is a pressing need for integrative studies that examine leaders, teams and their members, and contextual factors pertaining to the organization (Avolio, 2007; Graen & Ulh-Bien, 1995; Hackman & Wageman, 2007). 36 References Amabile, T. M. (1983). The social psychology of creativity: A componential conceptualization. Journal of Personality and Social Psychology, 45, 357–376. Amabile, T. M., Conti, R., Coon, H., Lazenby, J., & Herron, M. (1996). Assessing the work environment for creativity. Academy of Management Journal, 39, 1154–1184. Amabile, T. M. (1998). How to kill creativity. Harvard Business Review, 76, 77–87. Amabile, T. M., Schatzel, E. A., Moneta, G. B., Kramer, S. J. (2004). Leader behaviors and the work environment for creativity: Percieved leader support. The Leadership Quarterly, 15, 5–32. Anderson, N. R., & West, M. A. (1998). Measuring climate for work group innovation: Development and validation of the team climate inventory. Journal of Organizational Behavior, 19, 235–258. Anderson, N., de Dreu, C. K. W., & Nijstad, B. A. (2004). The routinization of innovation research: A constructively critical review of the state-of-the-science. Journal of Organizational Behavior, 25, 147–173. Aragón-Correra, J. A, García-Morales, V. J., & Cordón-Pozo, E. (2007). Leadership and organizational learning‘s role on innovation and performance: Lessons from Spain. Industrial Marketing Management, 36, 349–359. Archibugi, D., & Pianta, M. (1992). The technological specialization of advanced countries: A report to the ECC on international science and technology activities. Dordrecht: Kluwer Academic Publishers. Archibugi, D., & Pianta, M. (1996). Measuring technological change through patents and innovation surveys. Technovation, 16, 451–468. Atwater, L., & Carmeli, A. (2009). Leader–member exchange, feelings of energy, and involvement in creative work. The Leadership Quarterly, 20, 264–275. Avolio, B. J. (2007). Promoting more integrative strategies for leadership theory-building. American Psychologist, 62, 25–33. Avolio, B. J., Bass, B. M., & Jung, D. I. (1999). Re-examining the components of transformational and transactional leadership using the multifactor leadership questionnaire. Journal of Occupational and Organizational Psychology, 72, 441–462. Bain, P. G., Mann, L., & Pirola-Merlo, A. (2001). The innovation imperative: The relationships between team climate, innovation, and performance in research and development teams. Small Group Research, 32, 55–73. Baron, R. M., & Kenny, D. A. (1986). The moderator-mediator variable distinction in social 37 psychological research: Conceptual, strategic, and statistical considerations. Journal of Personality and Social Psychology, 51, 1173–1182. Barron, F., & Harrington, D. M. (1981). Creativity, intelligence, and personality. Annual Review of Psychology, 32, 439–476. Basadur, M. (2004). Leading others to think innovatively together: Creative leadership. The Leadership Quarterly, 15, 103–121. Basadur, M., Graen, G. B., & Scandura, T. A. (1986). Training effects on attitudes toward divergent thinking amongst manufacturing engieneers. Journal of Applied Psychology, 71, 612-617. Bass, B. M. (1985). Leadership and performance beyond expectations. New York: Free Press. Basu, R., & Green, S. G. (1997). Leader–member exchange and transformational leadership: An empirical examination of innovative behaviors in leader–member dyads. Journal of Applied Social Psychology, 27, 477–499. Burton, J. P., Sablynski, C. J., & Sekiguchi, T. (2008). Linking justice, performance, and citizenship via leader–member exchange. Journal of Business Psychology, 23, 51–61. Byrkjeflot, H. (2003). Nordic management: From functional socialism to shareholder value? In B. Czarniawska, & G. Sevón (Eds.), The Northern Lights: Organization theory in Scandinavia (pp. 17–39). Malmö, Sweden: Liber. Byrne, C. L., Mumford, M. D., Barrett, J. D., Vessey, W. B. (2009). Examining the leaders of creative efforts: What do they do, and what do they think about? Creativity and Innovation Management, 18, 256–268. Carlsson, B. (Ed). (1997). Technological systems and industrial dynamics. Dordrecht: Kluwer Academic Publishers. Csikszentmihalyi, M. (1999). Implications of a system perspective for the study of creativity. In R. J. Sternberg (Ed.), Handbook of creativity (pp. 313–334). Cambridge, UK: Cambridge University Press. Czarnitzki, D., & Kraft, K. (2004). Firm leadership and innovative performance: Evidence from seven EU countries. Small Business Economics, 22, 325–332. Damanpour, F. (1991). Organizational innovation: A meta-analysis of effects of determinants and moderators. Academy of Management Journal, 34, 555–590. Day, D. V., Schleicher, D. J., Unckless, A. L., & Hiller, N. J. (2002). Self-monitoring personality at work: A meta-analytic investigation of construct validity. Journal of Applied Psychology, 87, 390–401. 38 Deci, E. L., & Ryan, R. M. (1987). The support of autonomy and the control of behavior. Journal of Personality and Social Psychology, 53, 1024–1037. Denti, L., & Hemlin, S. (2011). What makes leadership and creativity connect? The mechanisms through which leaders may influence follower and team creativity. Manuscript. Gothenburg Research Institute, School of Business, Economics & Law and Department of Psychology, University of Gothenburg, Gothenburg, Sweden. Dougherty, D., & Hardy, C. (1996). Sustained product innovation in large, mature organizations: Overcoming innovation-to-organization problem. Academy of Management Journal, 39, 1120−1153. Drazin, R., Glynn, M. A., & Kazanjian, R. K. (1999). Multilevel theorizing about creativity in organizations: A sensemaking perspective. Academy of Management Review, 24, 286– 307. Dvir, T., Eden, D., Avolio, B. J., & Shamir, B. (2002). Impact of transformational leadership on follower development and performance: A field experiment. Academy of Management Journal, 45, 735–744. Eden, D. (1984). Self-fulfilling prophecy as a management tool: Harnessing Pygmalion. The Academy of Management Review, 9, 64–73. Edqvist, C. (Ed.). (1997). Systems of innovation: Technologies, institutions, and organizations. London: Pinter. Eisenbeiss, S. A., van Knippenberg, D., & Boerner, S. (2008). Transformational leadership and team innovation: Integrating team climate principles. Journal of Applied Psychology, 93, 1438–1446. Ekvall, G. (1996). Organizational climate for creativity and innovation. European Journal of Work and Organizational Psychology, 5, 105–123. Ekvall, G., & Ryhammar, L. (1999). The creative climate: Its determinants and effects at a Swedish university. Creativity Research Journal, 12, 303–310. Elenkov, D. S., & Manev, I. M. (2009). Senior expatriates leadership‘s effects on innovation and the role of cultural intelligence. Journal of World Business, 44, 357–369. Elkins, T., & Keller, R. T. (2003). Leadership in research and development organizations: A literature review and conceptual framework. The Leadership Quarterly 14, 587–606. Fai, F. (2007). A structural decomposition analysis of technological opportunity, corporate survival, and leadership. Industrial and Corporate Change, 16, 1069–1103. Feist, G. J., & Gorman, M. E. (1998). The psychology of science: Review and integration of a nascent discipline. Review of General Psychology, 2, 3–47. 39 Ford, C. M. (1996). A theory of individual creative action in multiple social domains. Academy of Management Review, 21, 1112–1142. Frese, M., Teng, E., & Wijnen, C. J. D. (1999). Helping to improve suggestion systems: Predictors of making suggestions in companies. Journal of Organizational Behavior, 20, 1139–1155. Frese, M., Fay, D., Hilburger, T., Leng, K., & Tag, A. (1997). The concept of personal initiative: Operationalization, reliability and validity in two German samples. Journal of Occupational and Organizational Psychology, 70, 139–161. Frohman, A. L. (1999). Personal initiative sparks innovation. Research-Technology Management, 42, 32–48. Gangestad, S. W., & Snyder, M. (2000). Self-monitoring: Appraisal and reappraisal. Psychological Bulletin, 126, 530–555. García-Morales, V. J., Mathías-Reche, F., & Hurtado-Torres, N. (2008). Influence of transformational leadership on organizational innovation and performance depending on the level of organizational learning in the pharmaceutical sector. Journal of Organizational Change Management, 21, 188–212. Garfield, E. (2006). The history and meaning of the journal impact factor. The Journal of the American Medical Association, 295, 90–93. George, J. M., & Zhou, J. (2001). When openness to experience and conscientiousness are related to creative behavior: An interactional approach. Journal of Applied Psychology, 86, 513– 524. Gerstner, C. R., & Day, D. V. (1994). Cross-cultural comparison of leadership prototypes. The Leadership Quarterly, 5, 121–134. Gerstner, C. R., & Day, D. V. (1997). Meta-analytic review of leader–member exchange theory: Correlates and construct issues. Journal of Applied Psychology, 82, 827–844. Gong, Y., Huang, J-C, & Fahr, J-L. (2009). Employee learning orientation, transformational leadership, and employee creativity: The mediating role of employee creative selfefficacy. Academy of Management Journal, 52, 765–778. Graen, G. B., & Cashman, J. F. (1975). A role-making model of leadership in formal organizations: A developmental approach. In J. G. Hunt and L. L. Larson (Eds.), Leadership frontiers. Kent, OH: Kent State University Press. Graen, G. B., & Uhl-Bien, M. (1995). Relationship-based approach to leadership: Development of leader–member exchange (LMX) theory of leadership over 25 years: Applying a multi-level multi-domain perspective. The Leadership Quarterly, 6, 219–247. Greguras, G. J., & Ford, J. M. (2006). An examination of the multidimensionality of supervisor 40 and subordinate perceptions of leader–member exchange. Journal of Occupational and Organizational Psychology, 79, 433–465. Gumusluoğlu, L., & Ilsev, A. (2009). Transformational leadership and organizational innovation: The roles of internal and external support for innovation. Journal of Product Innovation Management, 26, 264–277. Hackman, J. R., & Wageman, R. (2007). Asking the right questions about leadership. American Psychologist, 62, 43–47. Hemlin, S. (2006). Creative knowledge environments for research groups in biotechnology: The influence of leadership and organizational support in universities and business companies. Scientometrics, 67, 121–142. Hemlin, S. (2009). Creative knowledge environments: An interview study with group members and group leaders of university and industry R&D groups in biotechnology. Creativity and Innovation Management, 18, 278–285. Hemlin, S., Allwood, C. M., & Martin, B. R. (2008). Creative knowledge environments. Creativity Research Journal, 20, 196–210. Hemlin, S., Prpic, K., & Denti, L. (2009). Organizational support for innovation in R&D groups in biosciences: Comparing Sweden and Croatia. Manuscript. GRI/Department of Psychology, University of Gothenburg, Gothenburg, Sweden. Holmberg, I., & Åkerblom, S. (2006). Modelling leadership: Implicit leadership theories in Sweden. Scandinavian Journal of Management, 22, 307–329. Hooper, D. T., & Martin, R. (2008). Beyond personal leader–member exchange (LMX) quality: The effects of perceived LMX variability on employee reactions. The Leadership Quarterly, 19, 20–30. Howell, J. M., & Avolio, B. J. (1993). Transformational leadership, transactional leadership, locus of control, and support for innovation: Key predictors of consolidated-business-unit performance. Journal of Applied Psychology, 78, 891–902. Hülsheger, U. R., Anderson, N., & Salgado, J. F. (2009). Team-level predictors of innovation at work: A comprehensive meta-analysis spanning three decades of research. Journal of Applied Psychology, 94, 1128–1145. Hunter, S. T., Bedell, K. E., & Mumford, M. D. (2007). Climate for creativity: A quantitative review. Creativity Research Journal, 19, 69–90. Isaksen, S. G., Lauer, K. J., Ekvall, G., & Britz, A. (2001). Perceptions of the best and worst climates for creativity: Preliminary validation evidence for the situational outlook questionnaire. Creativity Research Journal, 13, 171–184. Jansen, J. J. P., Vera, D., & Crossan, M. (2009). Strategic leadership for exploration and 41 exploitation: The moderating role of environmental dynamism. The Leadership Quarterly, 20, 5–18. Jung, D. I. (2001). Transformational and transactional leadership and their effects on creativity in groups. Creativity Research Journal, 13, 185–195. Jung, D., Wu, A., & Chow, C. W. (2008). Towards understanding the direct and indirect effects of CEO‘s transformational leadership on firm innovation. The Leadership Quarterly, 19, 582–594. Kaiser, R. B., Hogan, R., & Craig, S. B. (2008). Leadership and the fate of organizations. American Psychologist, 63, 96–110. Kanter, R. M. (1996). When a thousand flowers bloom: Structural, collective, and social conditions for innovation in organizations. In P. S. Meyers (Ed.), Knowledge, management and organizational design (pp. 93–131). Boston, MA: ButterworthHeinemann. Keller, R. T. (2001). Cross-functional project groups in research and new product development: Diversity, communications, job stress, and outcomes. Academy of Management Journal, 44, 547–555. Kline, R. B. (2005). Principles and Practice of Structural Equation Modeling (2nd ed.). New York: Guilford Press. Krause, D. E. (2004). Influence-based leadership as a determinant of the inclination to innovate and of innovation-related behaviors: An empirical investigation. The Leadership Quarterly, 15, 79–102. Liden, R. C., & Maslyn, J. M. (1998). Multidimensionality of leader–member exchange: An empirical assessment through scale development. Journal of Management, 24, 43–72. Makri, M., & Scandura, T. A. (2010). Exploring the effects of creative CEO leadership on innovation in high-technology firms. The Leadership Quarterly, 21, 75–88. Mann, L. (2005). Leadership, management, and innovation in R&D project teams. London: Praeger. Marion, R., & Uhl-Bien, M. (2001). Leadership in complex organizations. The Leadership Quarterly, 12, 389–418. Miller, D., Dröge, C., & Toulouse, J-M. (1988). Strategic process and content as mediators between organizational context and structure. Academy of Management Journal, 32, 544– 569. Mumford, M. D., & Gustafson, S. B. (1988). Creativity syndrome: Integration, application, and innovation. Psychological Bulletin, 103, 27–43. 42 Mumford, M. D., Scott, G. M., Gaddis, B., & Strange, J. M. (2002). Leading creative people: Orchestrating expertise and relationships. The Leadership Quarterly, 13, 705–730. OECD. (1994). The measurement of scientific and technological activities: Using patent data as science and technology indicators. Paris: OECD. OECD. (2005). Oslo manual: Guidelines for collecting and interpreting innovation data (3rd ed.). Paris: OECD. Oldham, G. R., & Cummings, A. (1996). Employee creativity: Personal and contextual factors at work. Academy of Management Journal, 39, 607–634. Paulus, P. B., & Yang, H. C. (2000). Idea generation in groups: A basis for creativity in organizations. Organizational Behavior and Human Decision Processes, 82, 76–87. Pearce, C. L., & Ensley, M. D. (2004). A reciprocal and longitudinal investigation of the innovation process: The central role of shared vision in product and process innovation teams (PPITs). Journal of Organizational Behavior, 25, 259–278. Pelz, D., & Andrews, F. (1966). Scientists in organizations: Productive climates for research and development. New York: Wiley. Pierce, J. K., Gardner, D. G., Cummings, L. L., & Dunham, R. B. (1989). Organization-based self-esteem: Construct definition, measurement and validation. Academy of Management Journal, 32, 622–648. Pirola-Merlo, A. (2000). Innovation in R&D project teams: Modelling the effects of individual, team and organizational factors. Unpublished doctoral dissertation, University of Melbourne, Melbourne, Australia. Pirola-Merlo, A., & Mann, L. (2004). The relationship between individual creativity and team creativity: Aggregating across people and time. Journal of Organizational Behavior, 25, 235–257. Prabhu, V., Sutton, C., & Sauser, W. (2008). Creativity and certain personality traits: Understanding the mediating effect of intrinsic motivation. Creativity Research Journal, 20, 53–66. Puccio, G. J., Murdock, M. C., & Mance, M. (2007). Creative leadership: Skills that drive change. Thousand Oaks, CA: Sage. Rank, J., Nelson, N. E., Allen, T. D., & Xu, X. (2009). Leadership predictors of innovation and task performance: Subordinates‘ self-esteem and self-presentation as moderators. Journal of Occupational and Organizational Psychology, 82, 465–489. Rank, J., Pace, V. L., & Frese, M. (2004). Three avenues for future research on creativity, innovation, and initiative. Applied Psychology: An International Review, 53, 518–528. 43 Redmond, M. R., Mumford, M. D., & Teach, R. (1993). Putting creativity to work: Effects of leader behavior on subordinate creativity. Organizational Behavior and Human Decision Processes, 55, 120–151. Reiter-Palmon, R., & Illies, J. J. (2004). Leadership and creativity: Understanding leadership from a creative problem-solving perspective. The Leadership Quarterly, 15, 55–77. Scott, S. G., & Bruce, R. A. (1994). Determinants of innovative behavior: A path model of individual innovation in the workplace. Academy of Management, 37, 580–607. Seibert, S. E., Kraimer, M. K., & Crant, J. M. (2001). What do proactive people do? A longitudinal model linking proactive personality and career success. Personnel Psychology, 54, 845–874. Shalley, C. E., & Gilson, L. L. (2004). What leaders need to know: A review of social and contextual factors that can foster or hinder creativity. The Leadership Quarterly, 15, 33– 53. 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, 933–958. Shelley, C. E., & Perry-Smoth, J. E. (2000). Effects of social-psychological factors on creative performance: The role of informational and controlling expected evaluation and modelling experience. Organizational Behavior and Human Decision Processes, 84, 1–22. Sherony, K. M., & Green, S. G. (2002). Coworker exchange: Relationships between coworkers, leader member exchange, and work attitudes. Journal of Applied Psychology, 87, 542– 548. Shin, S. J., & Zhou, J. (2003). Transformational leadership, conservation, and creativity: Evidence from Korea. Academy of Management Journal, 46, 703–714. Simonton, D. K. (2003). Scientific creativity as constrained stochastic behavior: The integration of product, person, and process perspectives. Psychological Bulletin, 129, 475–494. Somech, A. (2006). The effects of leadership style and team process on performance and innovation in functionally heterogeneous teams. Journal of Management, 32, 132–157. Sosik, J. J., Avolio, B. J., & Kahai, S. S. (1997). Effects of leadership style and anonymity of group potency and effectiveness in a group decision support system environment. Journal of Applied Psychology, 82, 89–103. Sternberg, R. J. (Ed.). (1999). Handbook of creativity. Cambridge, UK: Cambridge University Press. Strannegård L., & Jönsson, S. (2009). Ledarskapets lockelse. In S. Jönsson & L. Strannegård (Eds.), Ledarskapsboken (pp. 11–27). Malmö, Sweden: Liber. 44 Tengblad, S. (2003). Den myndige medarbetaren. Strategier för ett konstruktivt medarbetarskap. Malmö, Sweden: Liber. Thomas, K. W., & Velthouse, B. A. (1990). Cognitive elements of empowerment: An ―interpretive‖ model of intrinsic task motivation. Academy of Management Review, 15, 666–681. Thompson, V. A. (1965). Bureaucracy and innovation. Administrative Science Quarterly, 10, 1– 20. Tidd, J., Bessant, P., & Pavitt, K. (2001). Managing innovation: Integrating technological, market and organizational change. Chichester, UK: Wiley. Tierney, P., & Farmer, S. M. (2004). The Pygmalion process and employee creativity. Journal of Management, 30, 413–432. 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. Trevelyan, R. (2001). The paradox of autonomy: A case of academic research scientists. Human Relations, 54, 495–525. Unsworth, K. L., Wall, T. D., & Carter, A. (2005). Creative requirement: A neglected construct in the study of employee creativity? Group & Organization Management, 30, 541–560. Wang, H., Law, K. S., Hackett, R. D., Wang, D., & Chen, Z. X. (2005). Leader–member exchange as a mediator of the relationship between transformational leadership and followers‘ performance and organizational citizenship behavior. Academy of Management Journal, 48, 420–432. West, M. A., & Anderson, N. R. (1996). Innovation in top management teams. Journal of Applied Psychology, 81, 680–693. West, M. A., Borril, C. S., Dawson, J. F., Brodbeck, F., Shapiro, D. A., & Haward, B. (2003). Leadership clarity and team innovation in health care. The Leadership Quarterly, 14, 393– 410. Williams, W. M., & Young, T. L. (1999). Organizational creativity. In R. J. Sternberg (Ed.), Handbook of creativity (pp. 373–391). Cambridge, UK: Cambridge University Press. Woodman, R. W., Sawyer, J. E., & Griffin, R. W. (1993). Toward a theory of organizational creativity. Academy of Management Review, 18, 293–321. Yuan, F., & Woodman, R., W. (2010). Innovative behavior in the workplace: The role of performance and image outcome expectations. Academy of Management Journal, 53, 323–342. 45 Leadership and innovation in organizations: A systematic review of factors that mediate or moderate the relationship a) Denti, L. Hemlin, S.b) a) Dept. of Psychology/Gothenburg Research Institute, University of Gothenburg. Email: leif.denti@gri.gu.se b) Dept. of Psychology/Gothenburg Research Institute, University of Gothenburg 1 Abstract A leader supports teams and individuals as they turn their creative efforts into innovations (leader as facilitator) and manages the organization‘s goals and activities aimed at innovation (leader as manager). This review focuses on when and how leadership relates to innovation (i.e., the factors that moderate or mediate the relationship between leadership and innovation). The sample consists of 27 empirical studies in which leadership is treated as the independent variable and innovation as the dependent variable. We used the following additional selection criteria in choosing the sample: the studies had to be published after 1980, indexed in the Web of Science, and published in journals with a Journal Impact Factor > 1.0. In addition to reviewing the moderating and mediating factors, we identified two factors where the findings are ambiguous. The review proposes five new factors that may mediate or moderate the relationship between leadership and innovation. Keywords: Innovation, creativity, LMX, transformational leadership, review 2 Fifty years ago Burns and Stalker (1961) published their influential work on management and innovation. Despite the significant body of innovation research since then, little attention has been paid to the role of leadership in innovation (Mumford, Scott, Gaddis, & Strange, 2002). One reason for this inattention may be the conviction that the non-linearity and complexity of the innovation process simply cannot be adequately and systematically managed (Tidd, Bessant, & Pavitt, 2001). In this explanation, leadership is only one of several factors that influence innovation in complex systems such as organizations (Kaiser, Hogan, & Craig, 2008) where there is considerable latitude for unpredictable dynamics (Marion & Uhl-Bien, 2001), including external forces and sometimes luck (Kaiser et al., 2008). However, as we demonstrate in this paper, other empirical research points in a different direction. This research shows that with their construction of environments where innovation and creativity can flourish, leaders are increasingly recognized as essential elements in their promotion (Hemlin, 2006a; Mumford et al., 2002). We view innovation in organizations as an outcome of individual, team, and organizational efforts joined to produce a new product, process, or service that is potentially attractive to a market. Innovation is then the result of a number of activities performed at different levels of the organization and in its external world. We find the following definition of innovation useful: "the implementation of a new or significantly improved product (good or service), or process, a new marketing method, or a new organizational method in business practices, workplace organization or external relations" (OECD, 2005, p. 46). 3 Sometimes innovation and creativity are used interchangeably in the literature (Basadur, 2004; Csikszentmihalyi, 1999). However, creativity is commonly viewed as idea generation (ideation) while implementation of ideas is innovation (Amabile, Conti, Coon, Lazenby, & Herron, 1996; Anderson, De Dreu, & Nijstad, 2004; Scott & Bruce, 1994). In this paper our focus is on innovation studies, but we also examine research that investigates innovation in terms of creativity when it is clear that innovation was the goal. Leadership and innovation Findings from recent empirical research suggest that leaders play an important role in advancing innovation in organizations and in connecting people in different organizations. Leaders perform this role in various ways. In the following section we look at two influential leadership theories that deal with innovation. Transformational and transactional leadership. It has been suggested that transformational leaders are well suited to the task of promoting creative and innovative goals (Bass & Riggio, 2006). Such leaders are concerned with elevating team members‘ goals, broadening their knowledge base, increasing their self-confidence, and providing motivation and opportunity for their personal development (Dvir, Eden, Avolio, & Shamir, 2002). Transformational leaders may champion creative ideas, act as role models (Howell & Higgins, 1990), and stimulate team members‘ motivation (Mumford et al., 2002). Transactional leaders, on the other hand, make exchanges (e.g., issue rewards or take corrective actions as the result of actions performed), set goals, and clarify objectives (Avolio, Bass, & Jung, 1999). On the individual and team levels, Eisenbeiss, van Knippenberg, and Boerner (2008) confirmed their hypothesis that transformational 4 leadership encourages team members‘ collaboration and results in increased development and implementation of ideas. In recent years researchers have begun to assess transformational leadership outcomes in studies of CEOs. García-Morales, Mathías-Reche, and Hurtado-Torres (2008) found a positive relationship between transformational leadership and innovation in their study of 408 CEOs in large Spanish companies who were asked to compare the general level of transformational leadership in their organizations with that in other organizations. Jung, Wu, and Chow‘s (2008) survey of upper level managers in 50 Taiwanese companies showed that the leadership style of CEOs was positively related to organizational innovative performance. Elenkov, Judge, and Wright (2005) confirmed this finding. A possible mechanism in the leader-innovation relationship may be the CEO-top management team (TMT) interface (Elenkov & Manev, 2005). The CEO may influence innovation and organizational entrepreneurship by encouraging team interaction, by decentralizing responsibilities and decision-making, and by influencing the TMT‘s propensity for risktaking (Ling, Simsek, Lubatkin, & Veiga, 2008). Leader-member exchange (LMX). LMX theory focuses on the dyadic relationship between leaders and team members (Gerstner & Day, 1997; Graen & Uhl-Bien, 1995). This theory, which assumes the leader develops a unique relationship with each team member, does not describe a general leadership style. Low quality LMX relationships are formal and impersonal since they are based primarily on employment contracts. In contrast, high quality LMX relationships extend beyond the formal contract because the leader and the team member exchange mutual trust, liking, respect, and influence. A few empirical studies 5 have been conducted on LMX and innovation. In their study of the quality of LMX relationships in a R&D facility in the U.S., Scott and Bruce (1994) found that LMX positively relates to team member innovative behavior. High LMX relationships may stimulate this behavior because of higher performance expectations and because team members may gain social face and status in the organization (Yuan & Woodman, 2010). Moreover, leaders in high quality LMX relationships may provide more opportunities for team member autonomy (Liden & Maslyn, 1998), leading to innovation (Hemlin, Allwood, & Martin, 2008). The dual process of managing innovation We believe that leadership is an integral part of innovative organizational performance for at least two reasons. First, leaders construct the environments that favor creativity and ultimately innovation (Shalley & Gilson, 2004). Much of the leadership research focuses on the essential leadership actions in this construction of context and opportunities that promote the bottom-up process of innovation. Leaders encourage intrinsic motivation (Avolio et al., 1999), facilitate problem solving (Tierney, Farmer, & Graen, 1999), foster a positive team climate (Anderson & West, 1998), and establish and maintain high quality work relationships with team members (Scott & Bruce, 1994). Second, in a top-down process, leaders manage the strategic innovation goals and activities of their organizations. Leaders may set these goals and direct these activities by managing time, facilities, money, and knowledge resources (Drazin, Glynn, & Kazanjian, 1999), by setting and managing individual and team goals, by defining expectations for creative performance (Shalley & Gilson, 2004), by managing rewards (Mumford & 6 Gustafson, 1988), and by granting autonomy to individuals and teams (Hemlin, 2006b; Hunter, Bedell, & Mumford, 2007). Thus, the leader orchestrates the dual process a) of providing support to teams and individuals as they turn their creative efforts into innovations (leader as facilitator) and b) of managing the organization‘s goals and activities aimed at innovation (leader as manager) (see Hemlin, 2006b). Scope of this study Although leaders have a significant impact on innovation activities, they don't work in a vacuum. Researchers have pointed to the power of the context, with its contingency factors, that may interact with leaders‘ efforts to stimulate and manage innovation (Bass & Riggio, 2006; Hunt & Conger, 1999; Mumford et al., 2002; Rank, Nelson, Allen, & Xu, 2009; Shalley & Gilson, 2004; Yukl, 1999). The contingency factors tell us when leadership relates to innovation (i.e., the circumstances), thus they moderate the relationship between leadership and innovation outcomes. In addition, we need more knowledge about the mechanisms leaders use to influence innovation. These are subsumed under another category of factors that mediate leadership and innovation, and they may tell us how leaders influence innovation (i.e., the leaders‘ influence at the individual, team, and organizational levels). This paper reviews 27 empirical studies that deal with the mechanisms by which leadership relates to innovation at the individual, team, and organizational levels of human behavior. Among others, Mumford et al., (2002), Oke, Munshi, and Walumbwa (2009), and Shalley and Gilson (2004) have addressed these factors in their research. However, there 7 are few systematic reviews of the empirical research that compile our current knowledge on the mechanisms and contingency factors (i.e., the mediating and moderating factors between leadership and innovation). For example, whereas Elkins and Keller (2003) studied only the effects of leadership on various outcomes in R&D, we take a broader approach. We also extend the work of Ford (1996) and Woodman, Sawyer, and Griffin (1993) by examining more recent research. Moreover, since several scholars have called for a better understanding of the relationship between leadership and innovation (e.g., Byrne, Mumford, Barrett, & Vessey, 2009; Mumford et al., 2002; Shalley & Gilson, 2004), we discuss five new, potential mediating and moderating factors on leadership and innovation. Furthermore, in addition to our recognition of the importance of these moderating and mediating factors, we acknowledge the importance of multiple levels of analysis where organizational processes are likely nested in different levels (Drazin et al., 1999; Ford, 1996; Hemlin et al., 2008). For example, the effect of leaders on employees may depend on the climate of the team and the culture of the organization. Our purpose in this paper is to analyze the relationship between leadership and innovation as reported in empirical studies with a focus on the factors that moderate and mediate these relationships. We begin by discussing previously suggested mediators and moderators from the literature. Then we suggest five new mediator and moderator mechanisms with empirically testable hypotheses. Procedure We conducted our literature search in several steps. We searched for journal articles using the online databases PsycINFO, PsycARTICLES, ISI Web of Science, Social 8 Services Abstracts, Sociological Abstracts, IBSS (International Bibliography of the Social Sciences), ASSIA (Applied Social Sciences Index and Abstracts), Business Source Premier, Econlit, and Regional Business News. We used the keywords leadership and innovation in our search. We also used the keyword creativity because it is occasionally used interchangeably with innovation in the literature. We analyzed each article‘s abstract in order to identify those articles that 1) were based on empirical studies and 2) treated leadership as an independent variable and innovation or creativity as a dependent variable. As we required that each journal article selected must have been peer reviewed, we did not consider unpublished articles and dissertations. At this point in our search we had identified 99 articles.1 In the next step, we eliminated articles from out list that were published before 1980 since we wanted to include only studies that a) used advanced methodologies not in significant use before 1980 and that b) reflected leaders‘ influence on employees with contemporary work attitudes and values. We then eliminated articles not indexed in the ISI Web of Science and articles with a journal impact factor > 1.0.2 We specified the latter two criteria in order to identify high quality journal articles that had gone through a more rigorous peer review process. Although questions have been raised about the JIF, (e.g.. Boor, 1982), it is generally recognized as a valid indicator of journal quality. When a journal‘s JIF is high, scholars are more inclined to submit their papers to that journal. More submissions increase rejection rates and provide a broader base for the selection of high 1 We were unable to locate two articles we identified as of interest. The journal impact factor (IF) of a journal reflects the average number of citations per article in the journal during the two preceding years. Thus when a journal has a IF of 1 or more, each article in the journal has been cited on average one time or more times in the two preceding years (Garfield, 2006). 2 9 quality research. JIFs have also been empirically related to external assessments of journal quality (Saha, Saint, & Christakis, 2003). Since we were interested only in analyzing the factors that mediate and moderate leadership and innovation, we coded the dependent variables (creativity or innovation) as either innovation-measures or creativity-only measures. The basis for this coding was our definition of innovation. The articles in our review had to measure some aspect of implementation (i.e., the application of ideas such as new products or processes). Thus, according to this criterion, we could include articles in which the research aimed at measuring creativity, but not articles in which the research measured creativity only. Results In our sample of 27 studies, 17 studies measured transformational/transactional leadership, 2 studies measured leader-member exchange (LMX), and 8 studies measured other leadership traits or behaviors. In the measurement of innovation, 10 of the 27 studies in the sample were conducted at the individual level, 3 studies at the team level, and 14 studies at the organizational level. Individual level mediating and moderating factors Creative self-efficacy, mediator. In an experimental study, Redmond, Mumford, and Teach (1993) manipulated participants‘ feelings of self-efficacy by asking leaders to tell participants whether their scores in a pre-test were well above average or average. The participants who were told they scored above-average exhibited greater confidence in solving the problems subsequently posed as well as greater creativity related to the products 10 produced. It has been suggested that high levels of self-efficacy may contribute to increased motivation (Ford, 1996), greater eagerness to pursue individual ideas, and more effective use of cognitive resources (Thomas & Velthouse, 1990). Gong, Huang, and Fahr (2009) recently tested this mediator and found that creative self-efficacy mediates the relationship between transformational leadership and employee innovative behavior. Organizational based self-esteem (OBSE), moderator. OBSE refers to the employee‘s self-perceived value as an organizational member (Pierce, Gardner, Cummings, & Dunham, 1989). Some empirical evidence suggests that employees with low OBSE benefit from transformational leadership since they doubt whether their ideas or efforts are of value to the organization. Rank et al. (2009) found that OBSE moderates the relationship between leadership and individual innovativeness since the relationship is more important to employees with low levels of OBSE. Self-presentation, moderator. Self-presentation, a core sub-dimension in the selfmonitoring construct, refers to the way in which individuals engage in impressionmanagement and modify their behavior in order to reflect the expectations from the social context (Gangestad & Snyder, 2000). High self-monitors tend to control and alter their behavior in order to present an image congruent with others‘ expectations. Such activity is less important for low self-monitors (Day, Schleicher, Unckless, & Hiller, 2002). Rank et al. (2009) confirmed their hypothesis that self-presentation moderates the relationship between transformational leadership and individual innovativeness. The relationship was stronger for employees with low self-presentation. Rank et al. concluded that low selfmonitors perform best when they are involved with an organization that supports them and 11 allows them to act independently. Therefore they perform best working under transformational leaders who recognize their individual strengths and needs. Team level mediating and moderating factors Team reflection, mediator. Team reflection is a communication undertaking where team members collectively consider the team's goals, strategies, and processes. Somech (2006) found that leaders influenced team reflection that in turn promoted team innovation. In functionally heterogeneous teams, a participative leadership style was necessary to promote team reflection; in homogeneous teams, a more direct style was appropriate. These results point to team reflection as a mediator for leadership and team innovation. Team heterogeneity, moderator. There are more opportunities for divergent thinking in heterogeneous teams where team members have diverse skill sets, knowledge, and cognitive problem solving styles (Hülsheger, Anderson, & Salgado, 2009; Keller, 2001). Somech (2006) examined how leadership behavior, operationalized as participative and directive, affected functional heterogeneity and innovation outcomes. In participative leadership, for example, there is shared influence in decision-making and solicitation of new ideas from team members. In directive leadership, for example, there is a framework for decision-making and clear rules for behavior. Somech found that participative leadership has a moderating effect on team innovation only when functional heterogeneity is high; directive leadership has an effect when team heterogeneity is low. Task characteristics, moderator. It is thought that complex tasks motivate employees to find creative ways to solve problems. With simpler tasks, employee fulfillment comes from solving problems using established knowledge and routines 12 (Shalley & Gilson, 2004). In Oldham and Cummings‘ (1996) study, although job complexity was not directly related to the production of patents, an interaction showed that high job complexity combined with high non-controlling and supportive supervision resulted in more patents compared to other combinations. Thus, the challenge of the task may moderate the effects of leadership on innovation. Organizational level moderating factors Organizational structure. Factors related to the structure of an organization are believed to influence innovation performance (Thompson, 1965). In their study of 50 Taiwanese firms, Jung et al. (2008) found that when centralization and formalization are low, the influence of transformational leadership on organizational innovation is greater. Organizations with a high degree of centralization typically have decision-making authority concentrated at the upper management level (Damanpour, 1991). High formalization in organizations, where numerous routines and rules regulate the work, is thought to impede organizational innovation (Damanpour, 1991). For example, high centralization and formalization may reduce the autonomy of creative employees and teams (Jung et al., 2008), may hamper inter-functional and inter-departmental collaboration (Miller, Dröge, & Toulouse, 1988), and may constrain employees who depart from established practices (Dougherty & Hardy, 1996). Organizational culture. The normative behavior expectations in an organization form the context in which individuals and teams are embedded (Hunter et al., 2007; Mumford et al., 2002). Researchers have recognized the importance of creating an organizational culture that encourages innovation (e.g., Shalley & Gilson, 2004). Other researchers have 13 shown that organizational support promotes innovative performance (e.g., Amabile et al., 1996; Pirola-Merlo & Mann, 2004). Risk-taking, experimentation, openness, trust, and autonomy are part of the organizational support that provides a foundation for innovation (Mumford et al., 2002). Recently, organizational support was examined in relation to leadership and innovation. Jung et al. (2008) found that organizational support (as defined by Siegel & Kaemmerer, 1978) moderated the CEO‘s leadership style and the firm‘s innovation. The extent of organizational support for innovation may influence the individual‘s willingness to undertake creative endeavors. Such willingness may in part depend on the perception of the consequences of such actions in the environment (Mumford & Gustafson, 1988; Shalley & Gilson, 2004). However, the findings are mixed regarding the role of organizational support. Scott and Bruce (1994) based their scale of organizational support on Siegel and Kaemmerer‘s (1978) study and found organizational support to mediate between LMX relationships and team members‘ innovative behaviors. Ambiguous contingency factors Under this heading, we present two factors where mixed findings point to the fact that it still is unclear whether they explain the relationship between leadership and innovation, i.e., if they act as mediating factors, or if they exert influence on leaders‘ discretion to lead innovative endeavors, i.e., if they act as moderating factors. Psychological empowerment. The first ambiguous factor is found at the individual level. Recent findings show that empowerment, the motivational construct consisting of meaning, competence, self-determination, and impact (Spreitzer, 1995), both moderates and 14 mediates the relationship between leadership and innovation. Pieterse, van Knippenberg, Schippers, and Stam (2010) found that, only for individuals with high levels of psychological empowerment, transformational leadership was positively related to innovative behavior and transactional leadership was negatively related to innovative behavior. However, Jung, Chow, and Wu (2003) and Jung et al. (2008) could not show that psychological empowerment at the organizational level moderated transformational leadership and organizational innovation. Furthermore, when creativity is the criterion variable, psychological empowerment has been shown to mediate between transformational leadership (Gumusluoğlu & Ilsev, 2009a) empowering leadership (Zhang & Bartol, 2010), and individual creativity. Thus, it is unclear whether leaders influence team members‘ psychological empowerment, as suggested by Gumusluoğlu and Ilsev (2009a), or if the construct is independent of leadership, as argued by Pieterse et al. (2010). Team climate. The second ambiguous factor is found at the team level. It is assumed that leaders generally have a significant influence in creating a climate conducive to team innovation by, for instance, acting as role models, supporting ideas, and participating in the work (Mumford et al., 2002). Consistent with these ideas, team climate appears to be a mediator between leadership and innovation as West et al. (2003) have shown. In their study, team climate, measured by Team Climate Inventory (TCI) (Anderson & West, 1998), partially mediated the effects of leadership influence on team innovation in selfmanaged teams. However, in a more recent study, Eisenbeiss et al. (2008) found that team climate moderated the positive relationship of transformational leadership and innovation Eisenbeiss et al. concluded that leaders may be an important part of team innovation, but without shared norms and ambitions of excellence, leaders‘ influence is limited. These 15 ambiguous findings point to a more profound problem in conceptualizing innovative climate. It is still unclear whether the leader primarily influences the team climate or is influenced by it. New moderators and mediators In this section, we propose five new mediators and moderators between leadership and innovation at the individual, team and organizational levels. We discuss their importance as mediators or moderators and propose a hypothesis for each. Some of these factors are known to the extant literature and have been tested as predictors for innovation. However, the factors proposed in this section have not yet been tested as mediators or moderators between leadership and innovation. Individual level External work contacts, mediator. The literature has emphasized the effect of external work contact networks on innovation (e.g., Tidd et al., 2001). This research has found that the number and frequency of such contacts relate to individual innovative work behavior (De Jong & den Hartog, 2010). De Jong (2007) hypothesized that the number of external contacts moderated the relationship between leadership and innovation, but was unsuccessful in finding evidence to support this relationship. However, Mumford et al. (2002) and Woodman et al. (1993) state that it is the job of leaders to organize and promote the information flow in groups. LMX theory stipulates that good leaders add their resources to those of their team members (Liden & Maslyn, 1998). Thus, the leader may influence 16 whether and when individuals seek contacts external to the group or organization. These external contacts may stimulate innovative endeavors (Hemlin & Olsson, 2011). Hypothesis 1. External work contacts for individuals mediate the relationship between leadership and innovation. Personal initiative, mediator. Personal initiative refers to the extent to which the individual engages in proactive and long-term, goal-oriented behaviors where actions are taken that extend beyond the terms specified in the formal work contract (Frese, Fay, Hilburger, Leng, & Tag, 1997). An individual‘s degree of personal initiative and proactiveness has been linked to innovation (Seibert, Kramier, & Grant, 2001). According to Frohman (1999), leaders may influence such behavior. Transformational leadership theory posits that transformational leaders motivate team members to make extracontractual efforts (Bass & Riggio, 2006). Similarly, a high quality LMX relationship may have this effect when a team member is given more responsibility and is placed in a position of greater trust (Liden & Maslyn, 1998). Hypothesis 2. Personal initiative is a mediator between leadership and innovation. Team level Group developmental stages, moderator. Various researchers have studied the different stages in group development (e.g., Arrow, Poole, Henry, Wheelan, & Moreland, 17 2004; Tuckman, 1965; Wheelan, 2005). In these group stages, cohesion, commitment, norm conformity, and goal related behavior may fluctuate (Tuckman, 1965). Leaders certainly influence the progression of these stages. Yet it appears that group members respond to leader influence differently in each developmental stage. In the integrative model of group development (Wheelan, 2005), leader influence depends on the group stage. In the first of four stages in this model, members tend to follow most of the direction provided by their leaders, but in the second stage, as they increasingly challenge this direction, conflict arises around issues such as roles, group organizing, and goals. In the third and fourth stages, when group members are again more open to leaders‘ influence, performance tends to be highest (e.g., Wheelan & Tilin, 1999). In this understanding of group development, the group dynamics involved in the different stages may limit leaders‘ influence to those processes involving innovation, such as ideation and idea implementation. Hypothesis 3. Group developmental stages moderate the relationship between leadership and innovation since the relationship will be strongest in the final stages. Creativity techniques for innovation, mediator. A leader may use a number of different creativity facilitating and creative problem solving techniques to enhance creativity and innovation in groups (Puccio, Murdock, & Mance, 2007; Rickards & Moger, 2006). McFadzean (1998) differentiates between three classes of creativity stimulating techniques: a) Those that make no attempt to change the existing paradigm, or frame of 18 thought; b) Those that stretch the paradigm; and c) Those that attempt to change the entire paradigm. For instance, leaders may initiate brainstorming sessions to encourage ideation, may appoint to a team member the role of the devil‘s advocate to promote idea reflection, or may employ guided imagination techniques to facilitate use of divergent information (Garfield, Taylor, Dennis, & Satzinger, 2001; McFadzean, 1998). Even if the leader does not use such techniques in ideation or in creative problem solving in the innovation process, the leader should initiate these activities. Hypothesis 4. Creativity techniques for innovation mediate the relationship between leadership and innovation. Organizational level Reward system, moderator. Employee motivation is frequently divided into intrinsic and extrinsic components. Amabile et al. (1996), among others, argue that the intrinsic components, i.e. the motivation that comes from approaching a complex task rather than external rewards related to its solution, are more related to creative and innovative efforts. However, some research indicates extrinsic components, such as contingent rewards (e.g., monetary incentives and recognition) for innovative pursuits, may also relate to creative performance (Eisenberger & Armeli, 1997). However, other findings are mixed since they show positive, negative, and null effects on creativity from such components (Baer, Oldham, & Cummings, 2003). To our knowledge, there is no research on organization rewards in conjunction with leadership and innovation. A leader may use the reward system to reward innovative efforts 19 by group members, but when a leader also rewards non-innovative efforts, group members may be less inclined to behave innovatively. In the latter case, leaders have greater difficulty in achieving innovative goals. Hypothesis 5. The reward system in an organization moderates the leadership and innovation relationship since the relationship strengthens when the reward system encourages innovative behavior. Conclusions The leadership and innovation link Based on this review of 27 peer-reviewed articles from the Web of Science‘s highest rated journals, we conclude that there is a relationship between leadership and innovation. Our review of the moderating and mediating factors on leadership and innovation (or creativity) reveals the importance of this relationship. We found the most frequently examined theory, transformational leadership (Bass, 1985), influences innovation at several levels of the organization, such as leader supervision of teams (Rank et al., 2009) and leader assumption of CEO functions (Elenkov et al., 2005). This conclusion agrees with our assumption that leaders have two roles in innovation: a) as facilitators for individuals and teams and b) as managers of organizational goals. Moreover, leaders who develop high quality LMX relationships with their employees promote innovative behavior through their emphasis on trust, respect, and contributions outside the terms of the formal work contract (Scott & Bruce, 1994). 20 Issues with the transformational leadership theory There are some issues about the transformational leadership construct, which was the most popular leadership theory in our review. These issues relate to how transactional leadership influences innovation. Conceptually, transformational leadership and transactional leadership are assumed to be separate leadership concepts rather than the opposite ends of a continuum of leadership styles. In their proposed model of leadership, Avolio and Bass (2002) argued that transformational leadership builds on transactional leadership. A fundamental principle of their model is that every leader exhibits both styles to some extent, described by Bass (1999) as the ―augmentation effect‖. In taking this view, we pose these (opposing) questions: (1) Does transformational leadership, in addition to transactional leadership, influence this outcome? (2) Does transformational leadership, unlike transactional leadership, stimulate innovation? While many studies focus only on the transformational dimension, a few studies examine both concepts. Rank et al. (2009) suggested that transactional leaders who closely monitor their subordinates are likely to suppress subordinates‘ intrinsic motivation and innovative behavior (see also Oldham & Cummings, 1996). Deci and Ryan (1987) found in a survey of 161 supervisor-employee pairs in 24 German companies that the dimension of transactional leadership is negatively related to employees‘ innovative behavior while transformational leadership is positively related. In a consideration of our two questions posed in the previous paragraph, we believe that transformational leadership in general has been inadequately measured in relation to 21 innovation. Scholars should measure both constructs, even when their intent is only to examine if transformational leadership, unlike transactional leadership, is related to innovation. We encourage the measurement of both constructs because a transformational leader also exhibits transactional behavior (Avolio & Bass, 2002; Bass, 1999). The When and the How: Contingency factors and mechanisms related to innovation In assessing these contingency factors related to when leaders may influence innovation, we conclude that the relationship between leadership and innovation appears strongest in organizations that have a supportive culture for innovation (Jung et al., 2008) and where organizational structures are de-formalized and de-centralized. In such organizations, both leaders and employees are freer to engage in creative work (Damanpour, 1991; Jung et al., 2008). Furthermore, teams that are heterogeneous and work on complex tasks have the highest capability for innovation. Such teams require supportive and non-controlling leadership that includes them in decision-making. Finally, leaders can promote innovative behavior among employees who have low organizational self-esteem and low self-presentation (Rank et al., 2009). In addressing the question of how leaders stimulate innovation (i.e., through the use of mediating variables) we conclude that leaders may stimulate innovation on the individual level by influencing creative self-efficacy (Gong et al., 2009; Redmond et al., 1993). Moreover, leaders may also stimulate innovation by introducing norms that encourage team reflection processes (e.g., by means of debates, open communication, and divergent thinking (Somech, 2006). 22 Future implications for leaders of innovation Hemlin et al. (2008) argued that a creative knowledge environment (CKE) should be established and promoted in organizations that wish to develop innovative products and processes. To a great extent, establishing a CKE is a leader responsibility. It is crucial for leaders to identify the specific environmental factors conducive to innovation and creativity. One may think of a CKE as a set of nested layers of environmental factors in an organization where individual and team creative activities are undertaken. In such environments, it is clear that the work design, as well as the social and organizational characteristics at the team and organizational levels, have a crucial influence on the innovation processes. This influence is implemented through adopting supportive cultures, informal structures, and organizational slack. Our literature research suggests there are a number of steps leaders may take when creating a CKE. First, upper management and their teams should establish an innovation policy that is promoted throughout the organization. It is necessary that the organization through its leaders communicate to employees that innovative behavior will be rewarded (Mumford & Gustafson, 1988). Second, when teams are composed, the potential for team innovation should be a consideration in selecting the team members. One team characteristic that seems to promote innovativeness is team heterogeneity (Reiter-Palmon & Illies, 2004). However, if the team is too heterogeneous, tensions may arise. On the other hand, when heterogeneity is too low, more directive leadership is required to promote team reflection, for example, by encouraging discussion and disagreement. Third, if creativity and innovativeness are to flourish, leaders should promote a team climate of emotional 23 safety, respect, and joy through emotional support and shared decision-making (Hemlin et al., 2008). Fourth, it is essential that individuals and teams have autonomy and space for idea generation and creative problem solving (Pelz & Andrews, 1966). Fifth, time limits for idea creation and problem solutions should be set, particularly in the implementation phases (Basadur, 2004; Mumford et al., 2002). Sixth, team leaders, who have the expertise, should engage closely in the evaluation of innovative activities (Mumford et al., 2002). Finally, leaders should adapt their cognitive and emotional support to others on an individual basis, e.g., a leader should assign tasks that are adapted to an individual team member‘s cognitive style and express personal trust to exploit innovative abilities of the person. Limitations of the study First, this paper is limited by its sampling procedures. We chose 1980 as our cut-off year because we wanted to include only research studies using advanced methodologies that were suitable for mediator-moderator analyses. We have no knowledge whether our results might have differed had we also included research studies published prior to 1980. Our findings may have also been affected by our requirement that the studies had to be of high quality (published in journals with JIF > 1.0). We omitted a few studies that tested mediator-moderator relationships but were published in lesser-ranked journals. Although we may also have omitted some recent high quality studies because they were published in dissertations or books, we considered this omission to be of less concern since it is likely those studies will be published at some future date in quality journals and will be included in future reviews. 24 Second, in the reviewed articles, theories and concepts were conceptualized, operationalized, measured and analyzed slightly differently, which then may affect the aggregated conclusions we reached. The risk is our conclusions may be too compartmentalized and/or oversimplified. This is an inherent problem of reviews of research literature. Suggestions for future research on leadership and innovation In addition to further research on the two ambiguous factors that we found (i.e., team climate and psychological empowerment), we suggest the following three areas for future research on leadership and innovation. Stages of the innovation process. Further research is needed into how the innovation process interacts with leaders‘ efforts. This process consists of problem construction or definition (Reiter-Palmon & Ilies, 2004), idea generation, evaluation, and promotion (Basadur, 2004), and of the planning, championing, and securing of funds for implementation (Scott & Bruce, 1994; Tushman & Nadler, 1986). The role of leaders is to provide a structure for the innovation process. In the early stages of innovation, leaders may have to take a divergent and explorative approach to problem construction and ideation in which knowledge and ideas are broadly integrated. Similarly, a convergent approach, focused on moving forward may be more suitable in the later stages where implementation is the focus. However, there is little research on how leaders may facilitate these cognitive and emotional processes in individuals and in teams. Negative leadership influence. Very few studies in our review deal with how leaders obstruct or impede innovation. For example, leaders who monitor their employees too 25 closely and give little support (Oldham & Cummings, 1996) who do not give them sufficient autonomy (Krause, 2004), who exclude them from the decision process, and who squelch new ideas (Somech, 2006) may stifle creativity and innovativeness. Therefore, research is needed on when and how leadership behaviors are detrimental to innovation. Leadership for radical and incremental innovation. More research is also needed on leadership when the goal is to create new and novel products (i.e., radical innovation) or when the goal is to expand and refine existing products (incremental innovation) (Tidd et al., 2001). While Jansen, Vera, and Crossan (2009) found a negative correlation between transactional leadership and radical innovation, their study showed that transactional leadership is positively related to incremental types of innovation. Thus, leaders who are more transaction-oriented may be more successful in promoting innovation when they work toward achieving the innovation goal by guiding employees towards refinements of existing products and increasing the efficiency of existing practices and processes. The distinction between incremental and radical innovation may thus be an important issue in future research and in theoretical modeling of leadership and innovation. Methodology. If contingency factors are addressed in studies in which leadership and management of innovation in organizations are examined, it is necessary to apply multi-level modeling and structural equation models. Such models are appropriate for analyzing the complex interrelationships of leadership and innovation. 26 References References marked with * were included in the selection of articles for this review. Amabile, T. M., Conti, R., Coon, H., Lazenby, J., & Herron, M. (1996). Assessing the work environment for creativity. Academy of Management Journal, 39, 1154-1184. Anderson, N., De Dreu, C. K. W., & Nijstad, B. A. (2004). The routinization of innovation research: A constructively critical review of the state-of-the-science. Journal of Organizational Behavior, 25, 147-173. Anderson, N. R., & West, M. A. (1998). Measuring climate for work group innovation: Development and validation of the team climate inventory. Journal of Organizational Behavior, 19, 235-258. *Aragón-Correra, J. A, García-Morales, V. J., & Cordón-Pozo, E. (2007). Leadership and organizational learning‘s role on innovation and performance: Lessons from Spain. Industrial Marketing Management, 36, 349-359. Arrow, H., Poole, M. S., Henry, K. B., Wheelan, S., & Moreland, R. (2004). Time, change and development: The temporal perspective on groups. Small Group Research, 35, 73-105. Avolio, B. J., & Bass, B. M. (2002). Developing potential across a full range of leadership. Cases on transactional and transformational leadership. Mahwah, NJ: Lawrence Erlbaum Associates. Avolio, B. J., Bass, B. M., & Jung, D. I. (1999). Re-examining the components of transformational and transactional leadership using the multifactor leadership questionnaire. Journal of Occupational and Organizational Psychology, 72, 441462. Baer, M., Oldham, G. R., & Cummings, A. (2003). Rewarding creativity: When does it really matter? The Leadership Quarterly, 14, 569-586. Basadur, M. (2004). Leading others to think innovatively together: Creative leadership. The Leadership Quarterly, 15, 103–121. Bass, B. M. (1985). Leadership and performance beyond expectations. New York: Free Press. Bass, B. M. (1999). Two decades of research and development in transformational leadership. European Journal of Work and Organizational Psychology, 8, 9-32. 27 Bass, B. M., & Riggio, R. E. (2006). Transformational Leadership (2nd ed.). Mahwah, NJ: Lawrence Erlbaum Associates. Boor, M. (1982). The citation impact factor: Another dubious index of journal quality. American Psychologist, 37, 975-977. Burns, T., & Stalker, G. (1961). The management of innovation. London: Tavistock. Byrne, C. L., Mumford, M. D., Barrett, J. D., & Vessey, W. B. (2009). Examining the leaders of creative efforts: What do they do, and what do they think about? Creativity and Innovation Management, 18, 256-268. *Chen, G., Tjosvold, D., & Liu, C. (2006). Cooperative goals, leader people and productivity values: Their contribution to top management teams in China. Journal of Management Studies, 43, 1177-1200. Csikszentmihaly, M. (1999). Implications of a system perspective for the study of creativity. In R. J. Sternberg (Ed.), Handbook of creativity (pp. 313-334). Cambridge, UK: Cambridge University Press. *Czarnitzki, D., & Kraft, K. (2004). Firm leadership and innovative performance: Evidence from seven EU countries. Small Business Economics, 22, 325-332. Damanpour, F. (1991). Organizational innovation: A meta-analysis of effects of determinants and moderators. Academy of Management Journal, 34, 555-590. Day, D. V., Schleicher, D. J., Unckless, A. L., & Hiller, N. J. (2002). Self-monitoring personality at work: A meta-analytic investigation of construct validity. Journal of Applied Psychology, 87, 390-401. Deci, E. L., & Ryan, R. M. (1987). The support of autonomy and the control of behavior. Journal of Personality and Social Psychology, 53, 1024-1037. De Jong, J. P. J. (2007). Individual innovation. The connection between leadership and employees’ innovative work behavior. Unpublished doctoral dissertation. Amsterdam: University of Amsterdam. De Jong, J., & den Hartog, D. (2010). Measuring innovative work behavior. Creativity and Innovation Management, 19, 23-36. Dougherty, D., & Hardy, C. (1996). Sustained product innovation in large, mature organizations: Overcoming innovation-to-organization problem. Academy of Management Journal, 39, 1120−1153. Drazin, R., Glynn, M. A., & Kazanjian, R. K. (1999). Multilevel theorizing about creativity in organizations: A sensemaking perspective. Academy of Management Review, 24, 28 286-307. Dvir, T., Eden, D., Avolio, B. J., & Shamir, B. (2002). Impact of transformational leadership on team member development and performance: A field experiment. Academy of Management Journal, 45, 735-744. *Eisenbeiss, S. A., van Knippenberg, D., & Boerner, S. (2008). Transformational leadership and team innovation: Integrating team climate principles. Journal of Applied Psychology, 93, 1438-1446. Eisenberger, R., & Armeli, S. (1997). Can salient reward increase creative performance without reducing intrinsic creative interest? Journal of Personality and Social Psychology, 72, 652-663. *Elenkov, D. S., Judge, W., & Wright, P. (2005). Strategic leadership and executive innovation influence: An international multi-cluster comparative study. Strategic Management Journal, 26, 665-682. *Elenkov, D. S., & Manev, I. M. (2005). Top management leadership and influence on innovation: The role of sociocultural context. Journal of Management, 31, 381-402. Elkins, T., & Keller, R. T. (2003). Leadership in research and development organizations: A literature review and conceptual framework. The Leadership Quarterly 14, 587606. Ford, C. M. (1996). A theory of individual creative action in multiple social domains. Academy of Management Review, 21, 1112-1142. Frese, M., Fay, D., Hilburger, T., Leng K., & Tag, A. (1997). The concept of personal initiative: Operationalization, reliability and validity in two German samples. Journal of Occupational and Organizational Psychology, 70, 139-161. Frohman, A. L. (1999). Personal initiative sparks innovation. Research-Technology Management, 42, 32-48. Gangestad, S. W., & Snyder, M. (2000). Self-monitoring: Appraisal and reappraisal. Psychological Bulletin, 126, 530-555. *García-Morales, V. J., Mathías-Reche, F., & Hurtado-Torres, N. (2008). Influence of transformational leadership on organizational innovation and performance depending on the level of organizational learning in the pharmaceutical sector. Journal of Organizational Change Management, 21, 188-212. Garfield, E. (2006). The history and meaning of the journal impact factor. The Journal of the American Medical Association, 295, 90-93. 29 Garfield, M., Taylor, N. J., Dennis, A. R., & Satzinger, J. W. (2001). Research report: Modifying paradigms – Individual differences, creativity techniques, and exposure to ideas in group idea creation. Information Systems Research, 12, 322-333. Gerstner, C. R., & Day, D. V. (1997). Meta-analytic review of leader-member exchange theory: Correlates and construct issues. Journal of Applied Psychology, 82, 827-844. *Gong, Y., Huang, J-C., & Fahr, J-L. (2009). Employee learning orientation, transformational leadership, and employee creativity: The mediating role of employee creative self-efficacy. Academy of Management Journal, 52, 765-778. Graen, G. B., & Uhl-Bien, M. (1995). Relationship-based approach to leadership: Development of leader-member exchange (LMX) theory of leadership over 25 years: Applying a multi-level multi-domain perspective. The Leadership Quarterly, 6, 219-247. *Gumusluoğlu, L., & Ilsev A. (2009a). Transformational leadership and organizational innovation: The roles of internal and external support for innovation. The Journal of Product Innovation Management, 26, 264-277. *Gumusluoğlu, L. & Ilsev, A. (2009b). Transformational leadership, creativity, and organizational innovation. Journal of Business Research, 62, 461-473. Hemlin, S. (2006a). Creative knowledge environments for research groups in biotechnology: The influence of leadership and organizational support in universities and business companies. Scientometrics, 67, 121-142. Hemlin, S. (2006b). Managing creativity in academic research. How could creative action and management be reconciled in research? Science Studies, 19, 3-12. Hemlin, S., Allwood, C. M., & Martin, B. R. (2008). Creative knowledge environments. Creativity Research Journal, 20, 196-210. Hemlin, S., & Olsson, L. (2011). Creativity-stimulating leadership: A critical incident study of leaders‘ influence on creativity in research groups. Creativity and Innovation Management, 20, 49-58. Howell, J. M., & Higgins, C. A. (1990). Champions of technological innovation. Administrative Science Quarterly, 35, 317-341. Hülsheger, U. R., Anderson, N., & Salgado, J. F. (2009). Team-level predictors of innovation at work: A comprehensive meta-analysis spanning three decades of research. Journal of Applied Psychology, 94, 1128-1145. Hunt, J. G., & Conger, J. A. (1999). From where we sit: An assessment of transformational and charismatic leadership research. The Leadership Quarterly, 10, 335-343. 30 Hunter, S. T., Bedell, K. E., & Mumford, M. D. (2007). Climate for creativity: A quantitative review. Creativity Research Journal, 19, 69-90. *Jansen, J. J. P., Vera, D., & Crossan, M. (2009). Strategic leadership for exploration and exploitation: The moderating role of environmental dynamism. The Leadership Quarterly, 20, 5-18. *Jung, D. I., Chow, C., & Wu, A. (2003). The role of transformational leadership in enhancing organizational innovation: Hypotheses and some preliminary findings. The Leadership Quarterly, 14, 525-544. *Jung, D., Wu, A., & Chow, C. W. (2008). Towards understanding the direct and indirect effects of CEOs‘ transformational leadership on firm innovation. The Leadership Quarterly, 19, 582-594. Kaiser, R. B., Hogan, R., & Craig, S. B. (2008). Leadership and the fate of organizations. American Psychologist, 63, 96-110. Keller, R. T. (2001). Cross-functional project groups in research and new product development: Diversity, communications, job stress, and outcomes. Academy of Management Journal, 44, 547-555. *Krause, D. E. (2004). Influence-based leadership as a determinant of the inclination to innovate and of innovation-related behaviors: An empirical investigation. The Leadership Quarterly, 15, 79-102. Liden, R. C., & Maslyn, J. M. (1998). Multidimensionality of leader-member exchange: An empirical assessment through scale development. Journal of Management, 24, 4372. *Ling, Y., Simsek, Z., Lubatkin, M. H., & Veiga, J. F. (2008). Transformational leadership's role in promoting corporate entrepreneurship: Examining the CEOTMT interface. Academy of Management Journal, 51, 557-576. *Makri, M., & Scandura, T. A. (2010). Exploring the effects of creative CEO leadership on innovation in high-technology firms. The Leadership Quarterly, 21, 75-88. Marion, R., & Uhl-Bien, M. (2001). Leadership in complex organizations. The Leadership Quarterly, 12, 389-418. McFadzean, E. (1998). The creativity continuum: Towards a classification of creative problem solving techniques. Creativity and Innovation Management, 7, 131-139. Miller, D., Dröge, C., & Toulouse, J-M. (1988). Strategic process and content as mediators between organizational context and structure. Academy of Management Journal, 32, 544-569. 31 Mumford, M. D., & Gustafson, S. B. (1988). Creativity syndrome: Integration, application, and innovation. Psychological Bulletin, 103, 27-43. Mumford, M. D., Scott, G. M., Gaddis, B., & Strange, J. M. (2002). Leading creative people: Orchestrating expertise and relationships. The Leadership Quarterly, 13, 705-730. OECD. (2005). Oslo manual. Guidelines for collecting and interpreting innovation data (3rd ed.). Paris: OECD. Oke, A., Munshi, N., & Walumba, F. O. (2009). The influence of leadership on innovation processes and activities. Organizational Dynamics, 38, 64-72. *Oldham, G. R., & Cummings, A. (1996). Employee creativity: Personal and contextual factors at work. Academy of Management Journal, 39, 607-634. *Osborn, R. N., & Marion, R. (2009). Contextual leadership, transformational leadership and the performance of international innovation seeking alliances. The Leadership Quarterly, 20, 191-206. Pelz, D., & Andrews, F. (1966). Scientists in organizations: Productive climates for research and development. New York: Wiley. Pierce, J. K., Gardner, D. G., Cummings, L. L., & Dunham, R. B. (1989). Organizationbased self-esteem: Construct definition, measurement and validation. Academy of Management Journal, 32, 622-648. *Pieterse, A. N., van Knippenberg, D., Schippers, M., & Stam, D. (2010). Transformational and transactional leadership and innovative behavior: The moderating role of psychological empowerment. Journal of Organizational Behavior, 31, 609-623. Pirola-Merlo, A., & Mann, L. (2004). The relationship between individual creativity and team creativity: Aggregating across people and time. Journal of Organizational Behavior, 25, 235-257. Puccio, G. J., Murdock, M. C., & Mance, M. (2007). Creative leadership: Skills that drive change. Thousand Oaks, CA: Sage. *Rank, J., Nelson, N. E., Allen, T. D., & Xu, X. (2009). Leadership predictors of innovation and task performance: Subordinates‘ self-esteem and self-presentation as moderators. Journal of Occupational and Organizational Psychology, 82, 45-489. *Redmond, M. R., Mumford, M. D., & Teach, R. (1993). Putting creativity to work: Effects of leader behavior on subordinate creativity. Organizational Behavior and Human Decision Processes, 55, 120-151. 32 Reiter-Palmon, R., & Illies, J. J. (2004). Leadership and creativity: Understanding leadership from a creative problem-solving perspective. The Leadership Quarterly, 15, 55-77. Rickards, T., & Moger, S. (2006). Creative leaders: A decade of contributions from creativity and innovation management journal. Creativity and Innovation Management, 15, 4-18. Saha, S., Saint, S., & Christakis, D. A. (2003). Impact factor: A valid measure of journal quality? Journal of the Medical Library Association, 91, 42-46. *Scott, S. G., & Bruce, R. A. (1994). Determinants of innovative behavior: A path model of individual innovation in the workplace. Academy of Management, 37, 580-607. Seibert, S. E., Kraimer, M. K., & Crant, J. M. (2001). What do proactive people do? A longitudinal model linking proactive personality and career success. Personnel Psychology, 54, 845-874. Shalley, C. E., & Gilson, L. L. (2004). What leaders need to know: A review of social and contextual factors that can foster or hinder creativity. The Leadership Quarterly, 15, 33-53. *Shin, S. J., & Zhou, J. (2003). Transformational leadership, conservation, and creativity: Evidence from Korea. Academy of Management Journal, 46, 703-714. Siegel, S. M., & Kaemmerer, W. F. (1978). Measuring the perceived support for innovation in organizations. Journal of Applied Psychology, 63, 553-562. *Somech, A. (2006). The effects of leadership style and team process on performance and innovation in functionally heterogeneous teams. Journal of Management, 32, 132157. Spreitzer, G. M. (1995). Psychological empowerment in the workplace: Dimensions, measurement and validation. Academy of Management Journal, 38, 1442-1465. Thomas, K. W., & Velthouse, B. A. (1990). Cognitive elements of empowerment: An ―interpretive‖ model of intrinsic task motivation. Academy of Management Review, 15, 666-681. Thompson, V. A. (1965). Bureaucracy and innovation. Administrative Science Quarterly, 10, 1-20. Tidd, J., Bessant, P., & Pavitt, K. (2001). Managing innovation. Integrating technological, market and organizational change. Chichester: Wiley. Tierney, P., Farmer, S. M., & Graen, G. B. (1999). An examination of leadership and 33 employee creativity: The relevance of traits and relationships. Personnel Psychology, 52, 591-620. Tuckman, B. W. (1965). Developmental sequence in small groups. Psychological Bulletin, 63, 384-399. Tushman, M., & Nadler, D. (1986). Organizing for innovation. California Management Review, 28, 74-92. *West, M. A., Borril, C. S., Dawson, J. F., Brodbeck, F., Shapiro, D. A., & Haward, B. (2003). Leadership clarity and team innovation in health care. The Leadership Quarterly, 14, 393-410. Wheelan, S. (2005). Group processes: A developmental perspective (2nd ed.). Boston: Allyn & Bacon. Wheelan, S., & Tilin, F. (1999). The relationship between faculty group development and school productivity. Small Group Research, 30, 59-81. Woodman, R. W., Sawyer, J. E., & Griffin, R. W. (1993). Toward a theory of organizational creativity. Academy of Management Review, 18, 293-321. *Yuan, F., & Woodman, R., W. (2010). Innovative behavior in the workplace: The role of performance and image outcome expectations. Academy of Management Journal, 53, 323-342. Yukl, G. (1999). An evaluation of conceptual weaknesses in transformational and charismatic leadership theories. The Leadership Quarterly, 10, 285-305. Zhang, X., & Bartol, K. M. (2010). Linking empowering leadership and employee creativity: The influence of psychological empowerment, intrinsic motivation, and creative process engagement. Academy of Management Journal, 53, 107-128. 34 Leader-member exchange and individual innovation: Intrinsic motivation and personal initiative as mediators, and the role of organizational support Denti, L. (a) and Hemlin, S. (b) (a) Dept. of Psychology / Gothenburg Research Institute, University of Gothenburg. Email: leif.denti@gri.gu.se (b) Dept. of Psychology / Gothenburg Research Institute, University of Gothenburg 1 Abstract This paper addresses the issue of whether individuals‘ characteristics or leaders‘ influence is more conductive to innovation in highly complex research and development (R&D) settings. The paper reports on a survey of 166 R&D team members, 43 team leaders, and 10 department managers in five Swedish industrial organizations. Innovative behavior was measured using quantitative indicators (e.g., new patents, new products, and scientific and other publications) and team leaders‘ ratings. The findings are that personal initiative predicted individual innovative behavior, while intrinsic motivation and leadership (conceptualized by leader-member exchange theory) did not. A mediating effect was found where leader-member exchange was associated with innovation through the personal initiative of team members. Organizational support moderated the relationship between leader-member exchange and individuals‘ personal initiative. High organizational support strengthened the relationship. Keywords: Leadership, innovation, creativity, intrinsic motivation, personal initiative 2 Innovation is an integral part of organizational performance. In recent decades the issue of how to promote innovation has become an intensive research area. Innovative outcomes are the result of the work of individuals. Therefore, without motivated and proactive employees, organizations stagnate and lose their competitive edge. For this reason, management in organizations tries to secure and support innovation. However, empirical research on innovation management in organizations is somewhat fragmented. Few attempts have been made to integrate the mechanisms that drive innovation at the level of work teams. In this paper we research the integration of organizational support, the management of R&D project teams, and individual efforts in an attempt to understand innovation at the individual and team levels in the industrial sector. This study addresses the most important antecedents of individual innovation. These factors are leaders‘ influence (Mumford, Scott, Gaddis, & Strange, 2002), the amount of organizational support (Amabile, Conti, Coon, Lazenby, & Herron, 1996; Ekvall & Ryhammar, 1999) and intrinsic motivation (Amabile, 1983; Woodman, Sawyer, & Griffin, 1993). We also address the more action-oriented behavioral construct of personal initiative that has been researched in performance studies (Frese, Fay, Hilburger, Leng, & Tag, 1997). By combining these factors in a single model, we were able to examine them as predictors, moderators, and mediators of individual innovation in the industrial research and development setting. 3 Individual innovation We agree with the following definition of innovation: "The implementation of a new or significantly improved product (good or service), or process, a new marketing method, or a new organizational method in business practices, workplace organization or external relations" (OECD, 2005, p. 46). Thus, individual innovation is the individual‘s active involvement in the development of new products, processes, or methods. Furthermore, while the various research and management disciplines sometimes use the concepts of innovation and creativity interchangeably (Csikszentmihalyi, 1999), creativity differs from innovation in that creativity refers to idea generation or ‗ideation‘ (e.g., Amabile et al., 1996), while innovation refers to the implementation of ideas (Anderson, De Dreu, & Nijstad, 2004). Thus, creativity precedes innovation in a multi-stage process where ideas are transformed into innovative outcomes (Scott & Bruce, 1994). Leader-member exchange (LMX) and individual innovation Leaders play a crucial role in facilitating and supporting innovation among members of their teams (Mumford et al., 2002; Scott & Bruce, 1994). In this paper, we claim that leadership is an interactional and inter-relational phenomenon since leaders influence team members in their work relationship. Leader-member-exchange (LMX) theory focuses on this dyadic relationship in which leaders engage in an exchange relationship with team members (Gerstner & Day, 1997; Graen & Uhl-bien, 1995). Various researchers have associated the quality of the LMX relationship with employee innovation (e.g., Basu & Green, 1997; Scott & Bruce, 1994; Yuan & Woodman, 2010). Low quality LMX relationships are based predominantly on formal work contracts, where, for instance, team 4 members exchange their labor for money and other benefits. In high quality LMX relationships, by contrast, both leaders and team members contribute more to the exchange than their terms of their work contracts require. Such relationships involve mutual trust, respect, liking, and joint contribution toward work goals (Greguras & Ford, 2006). LMX theory posits that in the development of work relationships, leaders and team members gradually enter into reciprocal exchanges of greater and greater value (Graen & Cashman, 1975). In the early stages of the relationship, a leader assesses the motivation, behavior, and performance of a team member in order to determine how much discretion, autonomy, and influence to give that team member (Graen & Cashman, 1975; Scott & Bruce, 1994). Researchers have associated discretion, autonomy, and influence, which are more common in high quality LMX relationships, with innovative behavior (e.g., Oldham & Cummings, 1996; Pelz & Andrews, 1966). Thus, we propose the following hypothesis: H1: The quality of the leader-member exchange relationship is positively associated with member innovation. The mediating role of employee intrinsic motivation Intrinsic motivation is a motivational state that is elicited by the characteristics and challenges of a task or problem rather than by incentives such as monetary rewards (Amabile, 1983). It is assumed that intrinsic motivation is among the most important factors that stimulate individual creativity (Amabile, 1983; Woodman et al., 1993). 5 Although the linkage is inconclusive, intrinsic motivation as an influential factor has been linked with individual creativity in conjunction with leadership (e.g., Dewett, 2007; Jaussi & Dionne, 2003; Shin & Zhou, 2003; Tierney, Farmer, & Graen, 1999). However, Gumusluoğlu and Ilsev (2009) were unable to show that intrinsic motivation mediated transformational leadership and individual creativity. To our knowledge, intrinsic motivation has not been tested with individual innovation as the outcome in conjunction with leadership although Tierney et al. (1999), in relating intrinsic motivation creativity to leadership, used so-called invention disclosures as one outcome measure. One aim of our study is to test if intrinsic motivation mediates leadership and innovation. Leaders may directly motivate their team members by pioneering a vision, by providing support and encouragement, and by paying individual attention to team members and their intellectual potential (Amabile et al., 1996; Avolio & Bass, 2002). Atwater and Carmeli (2009) stated that a high quality LMX relationship may increase the intrinsic motivation of employees by giving them autonomy and a sense of belonging and by recognizing their competences. Furthermore, leaders may stimulate intrinsic motivation in team members by setting goals that are difficult yet attainable (Locke, 1968; Locke, Latham, & Erez, 1988) and by giving more challenging structure to tasks (Amabile, 1998). In summary, since it is assumed that leaders influence employees‘ intrinsic motivation, leading to outcomes such as creativity, good performance and innovation, it is important to test these assumptions empirically to see if intrinsic motivation is a mediating 6 variable between leadership and innovation. Thus, we propose the following three hypotheses: H2a: The quality of the leader-member exchange relationship is positively related to members‘ intrinsic motivation. H2b: The degree of intrinsic motivation is positively related to members‘ innovation. H2c: Intrinsic motivation is a mediating variable between the leader-member exchange relationship and members‘ innovation. The mediating role of employee personal initiative Personal initiative has been defined as ―a behavior syndrome resulting in an individual‘s taking an active and self-starting approach to work and going beyond what is formally required in a given job‖ (Frese et al., 1997, p. 140). Although personal initiative is conceptually similar to intrinsic motivation, intrinsic motivation is a psychological state (Amabile, 1983) whereas personal initiative is a behavioral construct (Frese et al., 1997). Thus, personal initiative may be seen as the behavioral expression of an individual‘s motivational state. Individuals with high personal initiative are proactive and set goals that go beyond the terms of their formal work contracts. We think that personal initiative, in relation to innovation where the emphasis is on the implementation of ideas rather than ideation, is worth exploring because of its inherent behavioral- and action-oriented focus (Frese et al., 1997; Rank, Pace, & Frese, 2004). While there is as yet no empirical study 7 that has established a link between personal initiative and innovation, Seibert, Kraimer, and Crant (2001) found that the related concept of personal proactiveness, of which personal initiative is a sub-dimension, was positively related to innovation and to career success. In addition, Binnewies, Ohly, and Sonnentag (2007) found that personal initiative relates to creativity. In this study, we respond to Rank et al.‘s (2004) call to examine the role of personal initiative in innovation and to Frohman‘s (1999) suggestion that leaders may have a role in influencing team members‘ personal initiative. We hypothesize that personal initiative is a mediating factor between leadership and innovation. First, leaders may directly stimulate initiative, for instance, by recognizing employees‘ contributions, by maximizing top-down and lateral information conduits, and by postponing negative evaluations until after outcomes are shown to be inadequate (Frohman, 1999). The increased trust and mutual contribution in a close leader-member relationship may also stimulate team members‘ personal initiative. In such relationships, team members think leaders listen to their ideas and support their activities. Second, we claim that team members who are given positions of trust with increased autonomy—as is often the case in high quality LMX relationships (Liden & Maslyn, 1998)—may have more discretion to implement their ideas than team members who do not enjoy such trust and autonomy. Team members with strong personal initiative thrive in such trusting and independent relationships, and therefore they are more likely to satisfy the requirements set by the leader. Thus, we propose the following three hypotheses: 8 H3a: The quality of the leader-member exchange relationship is positively related to member personal initiative. H3b: Personal initiative is positively related to innovation. H3c: Personal initiative is a mediating variable between leader-member exchange and innovation. The moderating role of organizational support Leaders do not work in a vacuum. They lead team members who work in an organizational context commonly referred to as the organization‘s culture, climate, and practices (Hemlin, Allwood, & Martin, 2008). It is argued that organizations that support risk-taking, experimentation and promote trusting relationships provide a seedbed for innovation (Mumford et al., 2002). This support may lead to an increase in innovative performance (Ekvall & Ryhammar, 1999; Mann, 2005). In West and Anderson‘s (1996) study of hospital management groups, organizational support for innovation was the strongest predictor of overall innovation, defined as organizational change. Bain, Mann, and Pirola-Merlo (2001), who studied project teams in an R&D environment, confirm these findings. Mann (2005) concluded there are three forms of organizational support. The first form is organizational encouragement of innovation, which encompasses both the espoused value of innovation (i.e., the stated value of innovation) and the enacted value (i.e., the actual support for innovation). Mann‘s second form of organizational support is the 9 granting of resources. Resources include access to facilities, materials time, expert knowledge, and useful information. Mann‘s third form of organizational support is empowerment, which refers to employee autonomy (i.e., the freedom granted to pursue unique ideas and insights independently) and supervisory encouragement. Oldham and Cummings (1996) have linked both employee autonomy and supervisory encouragement to innovation. Ekvall (1996) and Hunter, Bedell, and Mumford (2007) have linked employee autonomy to innovation. Autonomy may be empowering for the group, sending signals of trust from the organization which invokes a sense of ownership and control (Amabile, 1998; Pirola-Merlo, 2000). We hypothesize that when an organization encourages innovation, provides resources, and empowers individuals, intrinsic motivation and personal initiative increase. Thus, we propose the following two hypotheses: Hypothesis 4a: The degree of perceived organizational support is positively related to intrinsic motivation. Hypothesis 4b: The degree of perceived organizational support is positively related to personal initiative. We also hypothesize that organizational support moderates the relationship between the leader‘s influence on team members‘ intrinsic motivation and personal initiative. When leaders work in a context where innovation is explicitly encouraged (i.e., a context in which sufficient resources are available and work group autonomy is permitted) it is more likely that they will provide such resources and grant such autonomy (Graen, Cashman, 10 Ginsburgh, & Schiemann, 1977). According to LMX theory, leaders and members continually make exchanges in order to achieve higher quality in the work relationship (Liden & Maslyn, 1998). Thus, when organizational support is high, leaders are in a better position to reward excellence and to build high quality work relationships that result in extra-contractual performance (Liden & Graen, 1980; Liden & Maslyn, 1998). When organizational support is weak, leaders have fewer opportunities to strengthen the work relationship with the members. Thus, the presence or absence of organizational support affects the ability of the leader to manage innovation among their members. Organizational support may thus act as a moderating variable between leadership and employee intrinsic motivation and personal initiative. Thus, we propose the following hypothesis: Hypothesis 4c. Organizational support for innovation will moderate the positive relationship between leader-member exchange and members‘ intrinsic motivation and personal initiative with the result that such relationships are stronger when organizational support is high. Purpose of the study The purpose of this study is to integrate and test several streams of research on factors related to individual innovation. Our hypothesis is that leadership, in the form of LMX, is mediated by team members‘ intrinsic motivation and personal initiative to achieve individual innovative outcomes. We also hypothesize that organizational support moderates 11 the relationship between LMX and team members‘ intrinsic motivation and personal initiative. Method Sample We surveyed 43 R&D teams from five industries (e.g., automotive, automation & control, industrial equipment, and paper manufacturing). Team members (chiefly scientists and engineers), team leaders, and department managers responded to the survey. Surveys that were sent to 211 team members were designed to measure the LMX relationship, organizational support, motivation, personal initiative and individual quantitative indicators of innovation. Surveys that were sent to the 43 team leaders were designed to measure the LMX relationship, quantitative indicators of innovation at the team level, and their ratings of individual innovative behavior. Surveys that were sent to the 15 department managers were designed to measure quantitative indicators of innovation at the team level and their ratings of team innovativeness. The response rates were as follows: team members, 78.6 percent (n = 166); team leaders, 100 percent (n = 43); and department managers, 66.7 percent (n = 10). Thus, the sample consisted of 166 leader-member dyads. The average age of the team members was 42 (SD = 10.7); 86 percent were men. As far as education, 11 percent of the team members had less than a bachelor‘s degree, 59 percent had a bachelor‘s or master‘s degree, and 30 percent had degrees higher than master‘s. The number of team members per team varied 12 from 4 to 50 (M = 13.8, SD = 8.5). On average, team members had worked in their teams for 4.1 years (SD = 9.0). Procedure In early 2010, we sent an invitation letter to 50 innovative organizations in Sweden. We selected these organizations based on the number of new patents issued by the Swedish patent and registration office. Three organizations responded to the letter and we contacted the CEOs who delegated the management of the survey participation to others. In two other organizations that were on the list of innovative organizations, we initiated contact with the R&D department managers directly. We asked all the R&D department managers to choose five team members at random. The teams had to have a clear innovation focus, such as research development or product development. We conducted our surveys on two occasions—Spring and Fall of 2010—with two reminders sent prior to each occasion. Measures Items in the survey were phrased as statements that participants rated on a sevenpoint Likert scale: (1 = Strongly disagree, 4 = Neither nor, 7 = Strongly agree). Unless otherwise specified, scales were calculated using the mean of items pertaining to the scale. Innovation. Past research has measured innovation variously. Researchers have used subjective ratings (e.g., Yuan & Woodman, 2010), experts‘ assessments (e.g., Jung, Wu, & Chow, 2008), and quantitative measures such as number of new patents (e.g., Jung et al., 2008), scientific publications (e.g., Tierney et al., 1999) and new products developed (e.g., 13 Garcia-Morales, Matias-Reche, & Hurtado-Torres, 2008). We also used these three quantitative measures plus the number of other publications (e.g., non-scientific reports, internal reports, folders, and promotional materials) (see Pirola-Merlo & Mann, 2005, who used other publications in his research). We asked the participants to report the number of 1) new patents, 2) scientific publications, 3) new products, and 4) other publications they had been involved in since joining the group under their current leader. In addition, we asked each team leader and each department manager to report the team‘s number of new patents, scientific publications, new products, and other publications during the same timeframe. We calculated the mean for the four measures. As this measure was severely skewed (skew = 2.86, SE = .19, kurtosis = 10.37, SE = .38), we transformed the mean into a natural logarithm with a constant (1) added to each measure in order to correct for nonnormality (Tabachnick & Fidell, 2006). We also used a six-item innovation rating scale developed by Scott and Bruce (1994). Using this scale, we asked each team leader to rate each team member‘s innovative behavior (α = .83) and each department manager to rate each team‘s innovativeness (α = .91). A sample item is: ―[Team member‘s name] searches out new technologies, processes, techniques, and/or product ideas‖. Table 1 shows that the team member quantitative innovation measure at the individual level significantly correlated with the individual subjective innovativeness ratings by team leaders (r = .194, p < .05). Since the team leaders and the department managers used the team as the measurement unit, we aggregated the individual ratings at 14 the team level. The team members‘ aggregated quantitative measure correlated with the team leaders‘ and department managers‘ quantitative innovation measures (r = .265, p < .01 and r = .335, p < .01, respectively). The correlation between the team leaders‘ and the department managers‘ quantitative innovation measures was also high and significant (r = .688, p < .01). When the team leaders‘ subjective assessment of innovativeness was aggregated at the team level, it correlated highly with the department managers‘ innovativeness ratings of the teams (r = .510, p < .01). Thus, using these three points of reference, we concluded that the dependent variable of innovation was measured adequately and with good convergent validity. In subsequent analyses, we used the individual level self-reported quantitative innovation measure and the individual level manager-rated subjective innovativeness scale (items 1 and 2 in Table 1). ___________________ Table 1 about here ___________________ Leadership. We measured leadership using the four dimensional, 12-item leadermember exchange measure (LMX-MDM) developed by Liden and Maslyn (1998). We asked team members and team leaders to rate each other on these four dimensions (α = .91 15 and α = .82, respectively). Using the electronic survey system, the participants could read either the team members‘ names or the team leaders‘ names instead of the original wording ―my supervisor‖. The four sub-dimensions in LMX-MDM are Contribution, Loyalty, Affect, and Professional Respect. These are explained next. Contribution: The amount, quality, and direction of mutual work-related activity aimed at achieving joint goals. A sample item is: ―I do work for [team leader‘s name] that goes beyond what is specified in my job description‖. The LMX-MDM scales for team leaders and team members were identical with the exception of this item. In the team leaders‘ version the item was: ―I provide support and resources for [team member‘s name] that goes beyond what is specified in my job description‖. Loyalty: The amount of mutual loyalty and obligation. A sample item is: ―[team leader‘s name] would defend me before others in the organization if I made an honest mistake‖. Affect: The amount of interpersonal liking. A sample item is: ―[team leader‘s name] is the kind of person one would like to have as a friend‖. Professional Respect: The respect for professional capabilities. A sample item is: ―I respect [leader‘s name]‘s knowledge of and competence on the job‖. The team member measure was subjected to a confirmatory factor analysis (CFA)1 where four latent variables were specified. The model fit indicated moderate but acceptable fit (χ2[48, n = 162] = 106.39, p < 0.001, χ2/df = 2.21; RMSEA = .087 [LO 90 = .065; HI 90 1 The software that was used here and in subsequent analyses is AMOS 18.0.0 build 992 (Arbuckle, 2009). 16 = .0109]; CFI = .96). The correlation between team member and team leader LMX measures was .115 (p > .15), which may be interpreted that the perceived agreement between the exchanges was low. In subsequent analyses, we used the team members‘ opinion of the leader-member relationship since this measure was theoretically more justified given the hypotheses of the study. Intrinsic motivation. We asked team members to rate their intrinsic motivation on a five-item scale developed by Tierney et al. (1999). A sample item is: ―I enjoy finding solutions to complex problems‖ (α = .72). Personal initiative. We used the six-item scale developed by Frese et al. (1997) to measure team members‘ personal initiative (α = .85). A sample item is: ―Whenever there is a chance to get actively involved, I take it‖. Organizational support. We used the 21-item OSIQ (Organizational Support for Innovation Questionnaire), developed by Pirola-Merlo (2000), consisting of three underlying constructs, to measure participants‘ perceptions of organizational support factors that facilitate innovation. The three constructs, Organizational encouragement of innovation, Access to resources, and Empowerment, are explained next. Organizational encouragement of innovation has sub-dimensions of i) espoused value of innovation, ii) enacted value of innovation, and iii) open discussion. A sample item is. ―This organization demonstrates its commitment to innovation by its decisions and 17 policies‖ (enacted value of innovation). Access to resources has sub-dimensions of i) access to experts, ii) facilities/materials, iii) information, and iv) time. A sample item is: ―In this organization there are usually knowledgeable people who can assist with difficult problems‖ (access to experts). Empowerment has sub-dimensions of i) autonomy and ii) supervisory encouragement. A sample item is: ―I am free to determine how I will allocate my time each day‖. The OSIQ scale was tested with CFA. Three latent factors were specified that corresponded to the three theoretical constructs of the scale. The errors of each manifest variable were allowed to correlate with other items within each sub-dimension of the three latent factors. Model fit indices indicated moderate fit (χ2[154, n = 162] = 296.25, p < 0.001, χ2/df = 1.92; RMSEA = .076 [LO 90 = .063; HI 90 = .089]; CFI = .91). Two items -- ―In this project the demand for facilities and materials generally exceeds the supply‖ and ―In this project, a hectic pace is required in order to meet time constraints‖ -- had low and non-significant regression values for their specified latent variables (b = .02, p > 0.81, and b = -.04, p > 0.61, respectively). A specification search on the standardized residual matrix (the difference between the observed covariance matrix and the matrix implied by the specified model) revealed several large, standardized residuals for these two items, indicating poor fit (> 3, values above 2.58 are considered to be large [Byrne, 2010]). In total, since the two items were deemed to have poor usability in the scale, we excluded them from further analyses. After they were excluded, the model fit was adequate: χ2(121, n = 162) = 229.18, p < 0.001, χ2/df = 1.89; RMSEA = .075 (LO 90 = .060; HI 90 = .089); CFI = .93. We excluded both items in the final scale of organizational support. In addition, we excluded two items related to 18 supervisory support from further analyses because of the risk of multi-collinearity of the independent variables when OSIQ is used in conjunction with the leadership measure and the risk of inflation of the interaction effect. The alpha for the organizational support scale was .90 (16 items). Methods of analysis To test for common method variance, we conducted Harman‘s one-factor test. This test is not a remedy for common method variance but is instead a statistical indicator of its severity (Podsakoff, MacKenzie, Lee, & Podsakoff, 2005). Because of the number of items, the dimensions of the multidimensional LMX and OSIQ scales were parceled. The test revealed five factors with an eigenvalue over 1 that explained 63.7 percent of the total variance. The first factor explained 26.6 percent of the variance. Each measure loaded on a corresponding factor, and the items of the intrinsic motivation measure loaded on two separate factors. Hence, we conclude that common method variance did not appear to be a severe problem. In accordance with Scott and Bruce‘s (1994) methodology, we tested our hypotheses using path-analysis. In path-analysis the researcher specifies a single model that can be used for a simultaneous analysis of the entire set of hypotheses. Path-analysis also allows the researcher to specify mediator and moderator mechanisms in the analysis of the relationship between independent and dependent variables. Before making our analyses, we screened for multivariate outliers using Mahalanobi‘s distance. We excluded four cases that were significant at the recommended p < 0.001 level (Kline, 2005). Multivariate normality was within the boundaries that Byrne (2010) recommends. To analyze moderation, we first 19 calculated z-scores to center the independent variables in order to calculate the interaction term that was the product of the LMX and OSIQ scales (Sauer & Dick, 1993). The interaction term, included in the path model as an exogenous variable, was used to predict the endogenous variables of intrinsic motivation and personal initiative. Similar to multiple regression analysis, significant parameter estimates of the paths would indicate a significant interaction effect (Baron & Kenny, 1986; Sauer & Dick, 1993). To test mediation, we used the product of coefficients test, which is considered superior in power and control of Type 1 errors to Baron and Kenny‘s (1986) traditional causal steps approach (MacKinnon, Lockwood, Hoffman, West, & Sheets, 2002). Maximum likelihood was used to estimate indirect paths as advocated by Brown (1997). To estimate the corresponding standard errors of the indirect paths, we used a bootstrapping method that is recommended when distributions are non-normal (Shrout & Bolger, 2002). This was the case with the dependent quantitative innovation measure in this study. In addition, we calculated 95 percent confidence intervals and p-values for the indirect paths using the bias corrected percentile method. 20 Results Table 2 presents descriptive statistics and inter-correlations for the two innovation measures and all predictors in the model. ___________________ Table 2 about here ___________________ Figure 1 presents results of the path analysis. The hypothesized model provided a good fit to the data when the dependent variable of innovation was the quantitative measure (Model 1) and when the dependent variable of innovation was the team leaders‘ ratings (Model 2). Model 1: χ2 (26, n = 162) = 29.84, p > .273, χ2/df = 1.50; RMSEA = .030 (LO 90 = .000; HI 90 = .072); CFI = .97. Model 2: χ2 (26, n = 156) = 30.02, p > .266, χ2/df = 1.16; RMSEA = .032 (LO 90 = .000; HI 90 = .074); CFI = .96. 21 ___________________ Figure 1 about here ___________________ The model accounted for 14 percent and 9 percent of the variances for the innovation measures of Model 1 and Model 2, respectively. Of the nine hypothesized paths, five paths were significant at the .05 alpha level. In the following analysis, the path estimates presented are for Model 1 where the quantitative innovation measure was the dependent variable (Model 1). LMX associated positively with intrinsic motivation and personal initiative in both models (Model 1: b = .22, p < .01 and b = .27, p < .01, respectively). This result supports Hypotheses 2a and 2b. However, only personal initiative was significantly related to team member innovation (Model 1: b = .22, p < .01). This result supports Hypothesis 3b. Contrary to Hypotheses 1 and 2b, the direct effect of LMX and intrinsic motivation was not significantly related to team member innovation. In Model 1, where employee innovation was the dependent variable, team member gender and the time a team member had been part of a team (group tenure) were significantly associated with team member innovation (b = .18, p < .05, and b = 15, p < .05, respectively). This result means that male team members and members that had been in the team longer reported more innovations. Additionally, organizational support was significantly related to intrinsic motivation (b = .17, p < .05), but not to personal initiative. This result supports Hypothesis 4a but not Hypothesis 4b. 22 Mediation analyses. Model 1 showed a significant indirect effect of LMX on team members‘ innovation with both estimation methods (i.e., the critical ratio test with bootstrapped standard errors (b = .081, SE = .041, CR = 1.98) and the bias corrected 95 percent confidence interval method (LO 95 = .021, HI 95 = .179, p < .01). In Model 2, the critical ratio of the parameter estimate and its bootstrap calculated standard error showed a tendency to be significant (b = .057, SE = .031, CR = 1.84), while the bias corrected confidence interval method was significant (LO 95 = .009, HI 95 = .131, p < .05). However, the above-stated effect represented the combined indirect effect of LMX on both intrinsic motivation and personal initiative of team members‘ innovation. Because intrinsic motivation and team member innovation were not significantly related, we re-ran both models excluding this variable to see if the indirect effect using personal initiative only was still significant. In both models and with both methods (i.e., the critical ratio test with bootstrapped standard errors, and the bias corrected confidence interval method), the indirect effect of LMX on team members‘ innovation by personal initiative was significant (Model 1: b = .058, SE = .028, CR = 2.07, LO 95 = .016, HI 95 = .131, p < .01; Model 2: b = .071, SE = .035, CR = 2.03, LO 95 = .020, HI 95 = .167, p < .01). Thus, Hypothesis 3c was supported while Hypothesis 2c was not. Moderation analyses. In both models, the interaction term was significantly associated with team members‘ personal initiative but not with their intrinsic motivation (Model 1: b = .20, p < .01, and b = 10, p > .18, respectively). Because the direct effect on personal initiative by OSIQ was not significant, we concluded that LMX interacts with OSIQ because higher levels of OSIQ create a stronger relationship between LMX and 23 personal initiative. Intrinsic motivation had positive relationships with OSIQ and LMX but we found no interaction. Thus, we concluded that Hypothesis 4c was partially supported. Discussion To our knowledge, this is the first study to test the individual level variables of intrinsic motivation and personal initiative as mediators in the hypothesized relationship between LMX and team members‘ innovation. We also hypothesized that LMX, intrinsic motivation, and personal initiative have a direct relationship with team members‘ innovation. Furthermore, the degree of support from the organization was hypothesized to moderate the relationship of LMX on team members‘ intrinsic motivation and personal initiative. Our hypothesis was that the relationship strengthened when support was high. We used two measures of innovation. The first was the combined number of new patents, products, scientific publications, and other publications that team members were involved with since working in the team under the current leader. The second measure of team member innovation was the team leaders‘ subjective ratings of their team members‘ innovative behaviors. Leader-member exchange in relation to intrinsic motivation and personal initiative The leader-member work relationship was positively related to team members‘ intrinsic motivation and personal initiative. This finding agrees with other studies (e.g., Atwater & Carmeli, 2009; Tierney et al., 1999). Thus, the emphasis on extra-contractual influences, such as trust, respect, and mutual contribution towards work goals (Greguras & 24 Ford, 2006), may influence team members‘ intrinsic motivation. In addition, to our knowledge our study is the first empirical study to link LMX to personal initiative. It has been theorized that in the negotiation of the LMX relationship, leaders reward such positive behaviors as initiative by giving members more autonomy, influence, and individual discretion in decision-making (Graen & Cashman, 1975; Scott & Bruce, 1994). We found that organizational support moderates the relationship of LMX on team members‘ personal initiative. When organizational support was greater, the relationship was strengthened. When team leaders and team members work in a context where innovation is actively encouraged (e.g., by providing expertise, information, facilities, and materials), team leaders have more flexibility in meeting their members‘ innovation needs. In an organization where autonomy is encouraged, team leaders may allow team members greater freedom and more discretion as far as decisions and actions (Frohman, 1999). Hence, as we hypothesized, where organizational support is high, team leaders can more easily build high quality work relationships with their team members. Such relationships may positively motivate team members to work innovatively. The role of personal initiative in innovative ventures When we used both quantitative measures of innovation and team leaders‘ ratings, we found that team members‘ personal initiative was directly related to their innovation. Contrary to our expectations, we found that team members‘ intrinsic motivation was not related to their innovation. While intrinsic motivation reflects an employee‘s mental state (Amabile, 1983), personal initiative is the behavioral enactment of intrinsic motivation 25 (Frese et al., 1997). Innovation work is often portrayed as a non-linear and risky process in which ideas often misfire, problems are ill-defined, and advances seldom follow an incremental path (Mumford et al., 2002). Moreover, innovation work involves more activities than idea generation. Such activities include idea development, championing of ideas, acquisition of resources, and taking steps toward implementation (Tidd, Bessant, & Pavitt, 2001). Based on the results of our study, we conclude that personal initiative may be a more pertinent construct for examining innovation than intrinsic motivation because it focuses on proactive behaviors that seem essential for innovation activities. Intrinsic motivation may not relate to team members‘ innovation in our study because the participants were already strongly and intrinsically motivated because of the inherently difficult challenge of their tasks (see Amabile, 1983) and of the high complexity of their own capabilities (see Feist & Gorman, 1998). Thus, the low variance in the intrinsic motivation variable in this sample, which consists mainly of scientists and engineers, may have produced the null relationship. Indeed, we also found that intrinsic motivation had the highest mean in combination with the lowest standard deviation of the variables predicting innovation. Leader-member exchange and team member innovativeness Particular attention should be paid to the result that we found no direct relationship between LMX and team members‘ innovation, either by our use of quantitative measures or by our use of team leaders‘ ratings. Thus, the work relationship between the team leaders and the team members does not appear to directly influence team members‘ innovation. This result contrasts with Scott and Bruce‘s (1994) and Yuan and Woodman‘s (2010) 26 studies in which they found positive, direct relationships in LMX with employee innovation as the outcome. However, we found that LMX was indirectly related to team member innovation through personal initiative when we used both the quantitative measures of innovation and the team leaders‘ ratings (i.e., personal initiative acted as a mediator between LMX and innovation). Our findings should be interpreted in the context of the sample. The sample consisted of team members who were mostly Swedish-based scientists and engineers, and of Swedish-educated team leaders. Swedish management principles have been contrasted with management principles in other cultural contexts (e.g., Byrkjeflot, 2003; Gerstner & Day, 1994; Holmberg & Åkerblom, 2006). To a great extent, in recent decades organizations in Sweden have experienced de-centralization where middle managers and their teams have assumed significant influence over organizational processes (Tengblad, 2003). In this management process, ideas such as egalitarianism, shared decision-making, employee collaboration, and team/individual autonomy have been emphasized (Holmberg & Åkerblom, 2006). Moreover, this process may have been accentuated in R&D departments. The results of this study, which focus on the characteristics of individuals and the supporting roles of leaders and organizations, may reflect Swedish organizational and managerial principles. Our results are also consistent with those in a recent study on Swedish industry- and university-based research groups that found team member-rated LMX unrelated to the publication of scientific research. Rather, the team leaders‘ ratings predicted this outcome (Olsson, Hemlin, & Pousette, 2010). Our results may also reflect the individual characteristics of the team members. In a meta-analysis of the personalities of scientists, Feist and Gorman (1998) concluded that 27 among other traits, scientists in general are more independent, driven, and achievementoriented than non-scientists. In addition, when highly creative scientists were compared with their less creative peers, this pattern of autonomy and achievement orientation was even more evident. Given the context of this sample it is possible that team leaders‘ influence may be a hygienic factor. That is, as long as the work relationship is reasonably satisfactory, the innovative work of team members will be adequately supported. However, a higher quality work relationship will not necessarily lead to more innovation. This conclusion agrees with Tierney et al. (1999) who found a difference between less and more innovative employees when leaders exert little influence over the latter group. Thus, we conclude that leaders may have a role to play in the facilitation of innovativeness, for example, by providing adequate resources and autonomy. Nevertheless, at least in the Swedish industrial environment, we claim that the personal characteristics of innovators are the more important factors behind innovation. Another finding of this study relates to the correlation between team leaders‘ and team members‘ perceptions of their work relationship. We found this relationship to be insignificant. This result is indicative of a well-known and serious problem in LMX theory. The problem may be caused by the LMX construct itself (see Greguras & Ford, 2006, on LMX-MDM), in the measurement of LMX in this study, or in the characteristics of the sample. In theory, LMX is a dyadic phenomenon (Schriesheim, Castro, Zhou, & Yammarino, 2001). Thus, the work relationship is viewed as an interactional, mutual exchange. Since both parts of the dyad should have a shared understanding of the 28 relationship, researchers have strongly encouraged measuring both parts (e.g., Schriesheim et al., 2001). However, research from the past 30 years shows that the covariance of the leaders‘ and member‘s ratings of the LMX relationship ranges only from 10 to 20 percent (Gerstner & Day, 1997). Our study is an improvement on past measurement of dyadic LMX because team leaders and team members could see their counterparts‘ names in the survey items. This procedure increased the clarity of their responses. Furthermore, we expected the leadermember agreement on LMX to be high since, according to tradition, Swedish management styles are thought to focus strongly on interpersonal relationships (Holmblad & Åkerblom, 2006). However, Olsson et al. (2010) have also found no significant correlation between leaders‘ and team members‘ LMX-MDM ratings in Swedish industrial or academic research groups. Thus, we question whether Greguras and Ford‘s (2006) use of LMX measures a dyadic phenomenon. Implications This study has implications for the management of innovation. First, department managers in human resources in their screening process for R&D new hires should look for industrial engineers who have initiative and motivational characteristics in addition to typical engineering skills. Secondly, team leadership should encourage and support innovation by giving team members‘ opportunities to use their individual initiative. Third, organizations should support innovation through adoption of policies that complement personal intrinsic motivation. They should adopt policies that give team members increased autonomy and that allocate sufficient resources to their projects. 29 Limitations and suggestions for further research We recognize there are limitations to this study. First, as we used a cross-sectional research design, our findings and inferences about relationships between variables cannot be interpreted as causal. It is possible that some relationships may be reciprocal. For instance, we hypothesized that team leaders influence the motivation and initiative of their team members. However, it is possible, and even likely, that team members who are already highly motivated affect the leader-member relationship positively. We suggest that future studies using longitudinal design take these kinds of reciprocal relationships into consideration. Second, as there are often highly complex relationships between the variables of organizational behavior, the relationships in our hypothesized model should be seen as one of several possible models where the variables differ in their juxtaposition to each other. For instance, it may be that organizational support mediates, rather than moderates, the relationship between LMX and intrinsic motivation and personal initiative. Since leaders often are the primary tools organizations use to support innovation, such support is created by leaders. Third, although we found a significant indirect path from LMX to individual innovation mediated through personal initiative, LMX had no direct relationship to innovation, a relationship that Baron and Kenny (1986) emphasize is necessary. We recognize more recent recommendations on mediation challenge this assumption of a direct relationship between the independent variable and the outcome (e.g., MacKinnon et al., 2002; Shrout & Bolger, 2002). Therefore, we acknowledge that the inference of mediation in this study may be prone to a Type 1 error. Fourth, as stated above, there was little agreement between team leaders and team members on the work 30 relationship. Since the relationship is theoretically conceptualized to take place at the dyadic level, further research is called for. We encourage researchers of meta-analyses of LMX theory to address this theoretical issue. Finally, since we obtained the predictors for innovativeness and the quantitative innovation measure only from team members, there is a risk of common source bias. Conclusions This study contributes to the vast area of leadership research by introducing a promising construct, personal initiative, in relationship to team members‘ innovation. Moreover, this study contributes to innovation management research. As innovative work generally is performed in unpredictable and complex environments (Kaiser, Hogan, & Craig, 2008; Marion & Uhl-Bien, 2001; Mumford et al., 2002), the ability to overcome obstacles proactively may be a key factor for in-house innovation. The amount of initiative shown by an individual may be a moderating factor in the relationship between creativity and innovation, since new ideas are more likely to be transformed into innovations when personal initiative is high (Rank et al., 2004). One strength of this study is that we tested the model with two different measures of innovation: quantitative measures and team leaders‘ ratings. These measures were further corroborated by three different sets of actors in the organizations. Such corroboration validates our findings and inferences about LMX, intrinsic motivation, and personal initiative in team members‘ innovation. 31 References Amabile, T. M. (1983). The social psychology of creativity: A componential conceptualization. Journal of Personality and Social Psychology, 45, 357-376. Amabile, T. M. (1998). How to kill creativity. Harvard Business Review, 76, 77-87. Amabile, T. M., Conti, R., Coon, H., Lazenby, J., & Herron, M. (1996). Assessing the work environment for creativity. Academy of Management Journal, 39, 1154-1184. Anderson, N., De Dreu, C. K. W., & Nijstad, B. A. (2004). The routinization of innovation research: A constructively critical review of the state-of-the-science. Journal of Organizational Behavior, 25, 147-173. Arbuckle, J. L. (2009). AMOS 18 User’s Guide. Crawfordville, FL: Amos Development Corporation. Atwater, L., & Carmeli, A. (2009). Leader-member exchange, feelings of energy, and involvement in creative work. The Leadership Quarterly, 20, 264-275. Avolio, B. J., & Bass, B. M. (2002). Developing potential across a full range of leadership. Cases on transactional and transformational leadership. Mahwah, NJ: Lawrence Erlbaum Associates. Bain, P. G., Mann, L., & Pirola-Merlo, A. (2001). The innovation imperative: The relationships between team climate, innovation, and performance in research and development teams. Small Group Research, 32, 55-73. Baron, R. M., & Kenny, D. A. (1986). The moderator-mediator variable distinction in social psychological research: Conceptual, strategic, and statistical considerations. Journal of Personality and Social Psychology, 51, 1173-1182. Basu, R., & Green, S. G. (1997). Leader-member exchange and transformational leadership: An empirical examination of innovative behaviors in leader-member dyads. Journal of Applied Social Psychology, 27, 477-499. Binnewies, C., Ohly, S., & Sonnentag, S. (2007). Taking personal initiative and communicating about ideas: What is important for the creative process and for idea creativity? European Journal of Work and Organizational Psychology, 16, 432-455. Brown, 1997. Assessing specific meditational effects in complex theoretical models. Structural Equation Modeling, 4, 142-156. Byrkjeflot, H. (2003). Nordic management: From functional socialism to shareholder value? In B. Czarniawska, & G. Sevón (Eds.) The Northern Lights – Organization Theory in Scandinavia (pp. 17-39). Malmö: Liber. 32 Byrne, B. M. (2010). Structural Equation Modeling with AMOS. Basic Concepts, Applications, and Programming (2nd ed). New York: Routledge. Csikszentmihaly, M. (1999). Implications of a system perspective for the study of creativity. In R. J. Sternberg (Ed.), Handbook of creativity (pp. 313-334). Cambridge, UK: Cambridge University Press. Dewett, T. (2007). Linking intrinsic motivation, risk taking, and employee creativity in an R&D environment. R&D Management, 37, 197-208. Ekvall, G. (1996). Organizational climate for creativity and innovation. European Journal of Work and Organizational Psychology, 5, 105-123. Ekvall, G., & Ryhammar, L. (1999). The creative climate: Its determinants and effects at a Swedish university. Creativity Research Journal, 12, 303-310. Feist, G. J., & Gorman, M. E., (1998). The psychology of science: Review and integration of a nascent discipline. Review of General Psychology, 2, 3-47. Frese, M., Fay, D., Hilburger, T., Leng, K., & Tag, A. (1997). The concept of personal initiative: Operationalization, reliability and validity in two German samples. Journal of Occupational and Organizational Psychology, 70, 139-161. Frohman, A. L. (1999). Personal initiative sparks innovation. Research-Technology Management, 42, 32-48. García-Morales, V. J., Mathías-Reche, F., & Hurtado-Torres, N. (2008). Influence of transformational leadership on organizational innovation and performance depending on the level of organizational learning in the pharmaceutical sector. Journal of Organizational Change Management, 21, 188-212. Gerstner, C. R., & Day, D. V. (1994). Cross-cultural comparison of leadership prototypes. The Leadership Quarterly, 5, 121-134. Gerstner, C. R., & Day, D. V. (1997). Meta-analytic review of leader-member exchange theory: Correlates and construct issues. Journal of Applied Psychology, 82, 827-844. Graen, G. B., & Cashman, J. F. (1975). A role-making model of leadership in formal organizations: A developmental approach. In J. G. Hunt and L. L. Larson (Eds.), Leadership Frontiers. Kent, OH: Kent State University Press. Graen, G. B., & Uhl-Bien, M. (1995). Relationship-based approach to leadership: Development of leader-member exchange (LMX) theory of leadership over 25 years: Applying a multi-level multi-domain perspective. The Leadership Quarterly, 6, 219-247. 33 Graen, G., Cashman, J. F., Ginsburgh, S., & Schiemann, W. (1977). Effects of linking-pin quality on the quality of working life of lower participants. Administrative Science Quarterly, 22, 491−504. Greguras, G. J., & Ford, J. M. (2006). An examination of the multidimensionality of supervisor and subordinate perceptions of leader-member exchange. Journal of Occupational and Organizational Psychology, 79, 433-465. Gumusluoğlu, L., & Ilsev, A. (2009). Transformational leadership and organizational innovation: The roles of internal and external support for innovation. The Journal of Product Innovation Management, 26, 264-277. Hemlin, S., Allwood, C. M., & Martin, B. R. (2008). Creative knowledge environments. Creativity Research Journal, 20, 196-210. Holmberg, I., & Åkerblom, S. (2006). Modelling leadership – implicit leadership theories in Sweden. Scandinavian Journal of Management, 22, 307-329. Hunter, S. T., Bedell, K. E., & Mumford, M. D. (2007). Climate for creativity: A quantitative review. Creativity Research Journal, 19, 69-90. Jaussi, K. S., & Dionne, S. D. (2003). Leading for creativity: The role of unconventional leader behavior. The Leadership Quarterly, 14, 475-498. Jung, D., Wu, A., & Chow, C. W. (2008). Towards understanding the direct and indirect effects of CEOs‘ transformational leadership on firm innovation. The Leadership Quarterly, 19, 582-594. Kaiser, R. B., Hogan, R., & Craig, S. B. (2008). Leadership and the fate of organizations. American Psychologist, 68, 96-110. Kline, R. B. (2005). Principles and Practice of Structural Equation Modeling (2nd ed.). New York: The Guilford Press. Liden, R. C., & Graen, G. (1980). Generalizability of the vertical dyad linkage model of leadership. Academy of Management Journal, 23, 451-465. Liden, R. C., & Maslyn, J. M. (1998). Multidimensionality of leader-member exchange: An empirical assessment through scale development. Journal of Management, 24, 4372. Locke, A. E. (1968). Toward a theory of task motivation and incentives. Organizational Behavior and Human Performance, 3, 157-189. Locke, A. E., Latham, P. G., & Erez, M. (1988). The determinants of goal commitment. Academy of Management Review, 13, 23-39. 34 MacKinnon, D. P., Lockwood, C. M., Hoffman, J. M., West, S. G., & Sheets, V. (2002). A comparison of methods to test mediation and other intervening variable effects. Psychological Methods, 7, 83-104. Marion, R., & Uhl-Bien, M. (2001). Leadership in complex organizations. The Leadership Quarterly, 12, 389-418. Mumford, M. D., Scott, G. M., Gaddis, B., & Strange, J. M. (2002). Leading creative people: Orchestrating expertise and relationships. The Leadership Quarterly, 13, 705-750. OECD. (2005). Oslo manual. Guidelines for collecting and interpreting innovation data (3rd ed.). Paris: OECD. Oldham, G. R., & Cummings, A. (1996). Employee creativity: Personal and contextual factors at work. Academy of Management Journal, 39, 607-634. Olsson, L., Hemlin, S., & Pousette, A. (2010). A multi-level analysis of leader-member exchange and creativity in research groups. Manuscript. GRI/Dept. of Psychology, University of Gothenburg. Pelz, D., & Andrews, F. (1966). Scientists in organizations: Productive climates for research and development. New York: Wiley. Pirola-Merlo, A. (2000). Innovation in R&D project teams: Modeling the effects of individual, team and organizational factors. Unpublished doctoral dissertation, The University of Melbourne: Melbourne. Pirola-Merlo, A., & Mann, L. (2005). Organizational supports for innovative R&D. In L. Mann (Ed.) Leadership, management, and innovation in R&D project teams. London: Praeger. Podsakoff, P. M., MacKenzie, S. B., Lee, J., & Podsakoff, N. P. (2003). Common method bias in behavioral research: A critical review of the literature and recommended remedies. Journal of Applied Psychology, 88, 879–903. Rank, J., Pace, V. L., & Frese, M. (2004). Three avenues for future research on creativity, innovation, and initiative. Applied Psychology: an International Review, 53, 518528. Sauer, P. L., & Dick A. (1993). Using moderator variables in structural equation models. Advances in Consumer Research, 20, 637-640. Schriesheim, C. A., Castro, S. L., Zhou, X., & Yammarino, C. C. (2001). The folly of theorizing ―A‖ but testing ―B‖. A selective level-of-analysis review of the field and a detailed leader-member exchange illustration. Leadership Quarterly, 12, 515-551. 35 Scott, S. G., & Bruce, R. A. (1994). Determinants of innovative behavior: A path model of individual innovation in the workplace. Academy of Management Journal, 37, 580607. Seibert, S. E., Kraimer, M. K., & Crant, J. M. (2001). What do proactive people do? A longitudinal model linking proactive personality and career success. Personnel Psychology, 54, 845-874. Shin, S. J., & Zhou, J. (2003). Transformational leadership, conservation, and creativity: Evidence from Korea. Academy of Management Journal, 46, 703-714. Shrout, P. E., & Bolger, N. (2002). Mediation in experimental and nonexperimental studies: New procedures and recommendations. Psychological Methods, 7, 422-445. Tabachnick, B. G., & Fidell, L. S. (2006). Using multivariate statistics (5th ed). Boston, MA: Allyn and Bacon. Tengblad, S. (2003). Den myndige medarbetaren. Strategier för ett konstruktivt medarbetarskap. Malmö: Liber. Tidd, J., Bessant, P., & Pavitt, K. (2001). Managing innovation. Integrating technological, market and organizational change. Chichester: Wiley. 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. West, M. A., & Anderson, N. R. (1996). Innovation in top management teams. Journal of Applied Psychology, 81, 680-693. Woodman, R. W., Sawyer, J. E., & Griffin, R. W. (1993). Toward a theory of organizational creativity. Academy of Management Review, 18, 293-321. Yuan, F., & Woodman, R., W. (2010). Innovative behavior in the workplace: The role of performance and image outcome expectations. Academy of Management Journal, 53, 323-342. 36 Tables 37 Table 1 Correlations between the quantitative innovation measures (logarithmized mean of patents, scientific publications, new products, and other publications (on the individual level and the team level), reported by team members, team leaders and department managers. Level 1 2 3 4 5 6 Individual 1. Team member quantitative measures (n = 166) Team 2. Leader subjective ratings (n = 161) .194* 3. Team member quantitative measures .662** .046 (aggregated) (n = 166) 4. Leader quantitative measures (n = 166) .178* .199* .265** 5. Department manager quantitative measures .228* .053 .335** .688** (n = 126) 6. Leader subjective ratings (aggregated) (n = -.026 .564** -.051 .205** .058 161) 7. Department manager subjective ratings (n = 126) *.Correlation is significant at the 0.05 level (2-tailed). ** Correlation is significant at the 0.01 level (2-tailed). 38 .025 .217* .026 .438** .510** .314** Table 2 Means, standard deviations, and inter-correlationsa between the study’s variables. Variable Mean SD 1 1. Employee quantitative measure 2.81 4.10 - 2. Leader subjective ratings 5.25 .83 .216** (.83) 3. Leader-member exchange quality 5.26 .93 .105 4. Intrinsic motivation 6.06 .67 .222** .121 5. Personal initiative 5.39 .73 .304** .239** .247** .457** (.84) 6. Organizational support 4.87 .79 .059 7. Genderb 8. Educationc 9. Group tenure (in years) 3.19 3.11 - 2 .079 .144 3 4 2.86 .233 ** 6 7 8 9 (.91) .242** (.71) .322** .240** .182* (.88) -.225** -.083 -.015 -071 1.26 .101 5 -.132 .051 - .114 .101 .159* .042 .183* -.022 .073 .058 -.012 .052 -.038 -.116 -.175* - 10. Age (in years) 42.10 10.30 -.075 .088 .020 -.025 -.036 -.014 -.049 -.143 .019 a n = 162, for variable 2, n = 157. Cronbach‘s coefficient alphas are given on the diagonal, where relevant. b Gender was coded as follows: 0 = ―male‖, 1 = ―female‖. c Education was coded as follows: 1 = ―Less than bachelor‘s degree‖; 2 = ―Bachelor‘s degree‖; 3 = ―Master‘s degree‖; 4 = ―Licenciate degree‖; 5 = ―Doctor‘s degree‖; 6 = ―Associate professor‖; 7 = ―Professor‖. *. Correlation is significant at the 0.05 level (two-tailed). **. Correlation is significant at the 0.01 level (two-tailed). 39 .17* (.17*) Organizational support (OSIQ) .09 (.08) .10 (.10) Intrinsic motivation .10 (.06) OSIQ x LMX ** * .20 (.20 ) Employee innovation .11 (.11) ** Personal initiative ** .22 (.21 ) LMX .14 (.09) .10 (-.01) .22** (.22*) .27** (.27**) .01 (.03) Gender Education Group tenure Age .18* (-.05) .09 (.14) .15* (.11) -.07 (.11) Figure 1a . Results for the hypothesized paths between leadership, and innovation outcomes mediated by intrinsic motivation and initiative and moderated by organizational support. a Two sets of parameter estimates are presented. The first set (Model 1) uses employee innovation as the dependent variable. The second set (Model 2) is in parentheses and uses team leaders‘ innovativeness ratings as the dependent variable. Standardized beta coefficients are given for the structural paths. Covariances are indicated with curves. R2 is presented for the endogenous variables. *. Significant at the 0.05 level (two-tailed). **. Significant at the 0.01 level (two-tailed). 40