Abstract
This article follows the call for more empirical research on leadership in different organizational contexts. Organizations react to environmental dynamism and uncertainty by using temporary forms or organizing such as projects. This organizational context factor raises questions about the nature of effective leadership in temporary settings. Transactional leadership has been found to be particularly effective in settings without a shared history of leaders and followers. In turn, prior research indicates that transformational leadership succeeds in times of uncertainty, which is essential to temporary organizations. We extend the existing transactional and transformational leadership approaches by examining leadership in the context of the temporary organization. We empirically test for effects of transactional and transformational leadership in projects and find that both leadership behaviors positively influence the followers’ commitment. However, transformational leadership is more effective than transactional leadership. The effects of both leadership behaviors are amplified by increasing complexity of the project.
Keywords
Introduction
Leadership research has accentuated the important role of the organizational context for leadership and has called for more empirical studies on leadership in different contexts (Antonakis, Avolio, & Sivasubramaniam, 2003; Avolio, Walumbwa, & Weber, 2009; Porter & MacLaughlin, 2006). We follow this call and present a study on the effects of transactional and transformational leadership on followers’ commitment in the context of the temporary organization. This form of organizing is increasing as organizations face growing uncertainties in dynamic environments (Bakker, 2010; Schreyögg & Sydow, 2010). Temporary organizations, such as programs or projects, are chosen for carrying out complex, novel, and knowledge-intensive tasks (Hobday, 2000; Lundin & Midler, 1998). Temporary forms of organizing are different from standard organizational processes as they are unique in terms of tasks and have a limited duration and a short-term orientation. Owing to this, temporary organizations are characterized by discontinuous personal constellations and work contents, a lack of organizational routines, and a cross-disciplinary integration of internal and external experts. In many cases, projects are carried out beyond hierarchical lines of authority and cut across organizational boundaries (Engwall, 2003; Hanisch & Wald, 2011). The characteristics of temporary organizations constitute a specific organizational context which is different from that of permanent organizations. As this context has become increasingly prevalent, its characteristics have become particularly relevant for examinations of the antecedents and coherence of leadership (Gundersen, Hellesøy, & Raeder, 2012).
Leadership theory has considered transactional and transformational leadership as two complementary points of view. Transactional leadership focuses on the task-related exchange of actions and rewards between followers and leaders, which often needs the existence of hierarchy and authority to be displayed. Transformational leadership sets in if leadership instruments such as awards and punishments, do not work. The leadership behavior associated with transformational leadership augments the active form of transactional leadership, that is, contingent reward. By aligning the followers’ needs with the organization’s (higher) tasks and goals (Bass, 1990), this person-oriented leadership style has been found to be especially effective in times of uncertainty and change (Waldman, Ramírez, House, & Puranam, 2001). Although empirical studies on transformational leadership have revealed important insights on its positive effects on followers’ commitment to change, these studies have predominantly analyzed both transformational and transactional leadership within the context of permanent organizations (Antonakis et al., 2003). Thus, transformational leadership rests on the assumption that relatively stable leader–follower constellations are prevalent. This assumption is not only reflected by the notion of the “emergence” of transformational leadership (Balthazard, Waldman, & Warren, 2009) but is also suggested in general discussions on different temporal sequences in leadership research (Bluedorn & Jaussi, 2008). Both leadership styles assume a common history and future, which are not necessarily given in temporary organizations (Bakker & Janowicz-Panjaitan, 2009).
The aim of this study is to elucidate the effects of leadership on employees’ commitment in the context of temporary organizations. Our research contributes to the literature in three ways. First, from the perspective of temporary organizations, we examine the effects of transactional and transformational leadership by considering each temporary organization’s particular characteristics. Second, this study offers insights on the challenges that leaders face in the context of temporary organizations. Because the underlying rationale for forming temporary organizations is that complex and extraordinary tasks can be performed through projects (Brockhoff, 2006; Clegg & Courpasson, 2004), we study the impact of this complexity on leadership effectiveness in projects. Third, we test our hypotheses using data covering different project and industry types and, therefore, generating broad empirical results.
Although effective leadership is vital to the success of all organizations (Kouzes & Posner, 2008), it is particularly important to organizations facing a high degree of uncertainty and change (Bass, 1990), as leadership positively affects employee commitment to change (Waldman et al., 2001). In turn, commitment is vital to meeting the objectives of temporary organizations (Rossy & Archibald, 1992). Transformational leadership is reported to have a positive effect on followers’ commitment and on performance (Avolio, Weichun, Koh, & Bhatia, 2004; McColl-Kennedy & Anderson, 2005; G. Wang, Oh, Courtright, & Colbert, 2011). We argue that transformational leadership is especially effective at establishing commitment to temporary organizations in the midst of uncertainty and change. We also investigate the success of transactional leadership in temporary organizations, as previous research has found such leadership to be especially important in situations marked by stress and weak social relations (Hannah, Uhl-Bien, Avolio, & Cavarretta, 2009).
Theory and Hypotheses
Characteristics of Temporary Organizations and Their Effects on Leadership
Temporary organizations are aggregates of individuals temporarily collaborating for a shared cause (Packendorff, 1995). Most temporary organizations are based on and set up by a permanent organization (Ekstedt, Söderholm, & Wirdenius, 1999) and take the form of projects, programs, temporary teams, or task forces. In the following, we use the terms temporary organization and project interchangeably as our empirical investigation focuses on projects as one important form of temporary organizations. Projects and other forms of temporary organizations are not solely created when extraordinary undertakings emerge but rather represent an increasingly large share of the ordinary operations of organizations (Engwall, 2003). Accordingly, there is a shift in perspective: Projects are no longer considered as exceptional cases but as regular business processes creating value (Engwall, 2003; Jalkala, Cova, Salle, & Salminen, 2010; Sydow, Lindkvist, & DeFillippi, 2004; Whitley, 2006). What distinguishes temporary teams from nontemporary teams is that they carry out time-limited undertakings and disperse on completion (Savelsbergh, Gevers, van der Heijden, & Poell, 2012). The relevance of temporary organizations as a field of study is not only reflected by an increase of “pure” forms such as projects and programs but also by the more general development of organizing in a more flexible and adaptable way. The ongoing discussion in organization theory on fluid organizations (Schreyögg & Sydow, 2010), duality (Farjoun, 2010), and organizational ambidexterity (Raisch, Birkinshaw, Probst, & Tushman, 2009), points to a development where permanent organizations increasingly develop and integrate temporary elements for either adapting to dynamic environments or for increasing innovativeness.
The characteristics of temporary organizations pose specific challenges to leadership (Chen, Donahue, & Klimoski, 2004). The operations of temporary organizations are time-limited (Maaninen-Olsson & Müllern, 2009) and unique in setup, tasks, and work content (Brockhoff, 2006). In contrast to standard organizational processes, temporary organizations draw on nonroutine processes that can cause uncertainty (Pich, Loch, & de Meyer, 2002). This uncertaintyisexacerbatedby the collaborations among people with heterogeneous backgrounds (Chen et al., 2004; Hoegl & Weinkauf, 2005) and a discontinuity of work constellations (Parker & Skitmore, 2005). Moreover, project managers often face an “authority gap” as project team members have differing (hierarchical) roles outside the temporary organization. It is usually the superior in the permanent organization and not the project manager who decides on promotions (Baccarini, 1996; Packendorff, 1995).
As a result of the characteristics of temporary organizations, project team members face a high degree of complexity, uncertainty, and ambiguity and are often less committed to their tasks than the employees of permanent organizations (Keegan & Den Hartog, 2004). Table 1 summarizes the characteristics of temporary organizations and the effects of these characteristics on the conditions of work in these organizations.
Characteristics of Temporary Organizations and Their Implications.
Prior research has shown that organizations must establish the employees’ commitment to the project’s objectives to ensure a strong overall team performance (Hoegl, Weinkauf, & Gemuenden, 2004) and the realization of the projects’ goals (Gällstedt, 2003). Therefore, leadership must positively influence each participant’s commitment to the project and inspire the firm’s employees to become a part of the project team (McDonough III, 2000). Accordingly, Zimmerer and Yasin (1998) identify poor or misdirected leadership to be one of the most important causes of project failure.
Transactional Leadership and Temporary Organizations
Transactional leadership primarily consists of two leadership principles: Leaders must award or reprove of their followers’ actions, and they must pay attention to all deviations and irregularities (Waldman et al., 2001). Transactional leaders reward their followers if they do what is expected of them and penalize them if they deviate from expectations (Bass, 1990). Because transactional leadership handles the day-to-day tasks, it constitutes the foundation of effective leadership behavior in organizations (Bass, Avolio, Jung, & Berson, 2003). In the context of temporary organizations, transactional leadership includes active monitoring of the followers’ actions to ensure that the projects are realized. Because the routines, processes, and structures related to permanent organizations are missing in temporary organizations (Kerzner, 2009), project management standards serve as substitutes for tracking the followers’ accomplishments. Prior research has shown that a structured course of action that uses standardized processes can positively influence the temporary organization’s capacity to achieve its targets (Pinto & Covin, 1989). A clearly defined project goal accompanied by a detailed project plan with delineated responsibilities allows for the pursuit of eventual deviations. Referring to Selznick (1957), Pawar and Eastman (1997) claim that leaders in times of relative goal stability resemble administrative managers rather than visionary leaders, who represent the transformational leadership style. Because a project’s short-term orientation and focus on immediate deliverables shorten each team member’s time horizon (Marks, Mathieu, & Zaccaro, 2001), scholars have claimed that project leaders should focus on the actual tasks to be accomplished, as time is too scarce to build relationships (Müller & Turner, 2007). If the desired performance is attained, then the reward for a follower does not require a deep relationship between the leader and the follower but can be displayed within time constraints (Den Hartog, van Muijen, & Koopman, 1997) and relationships that lack trust. With respect to temporary organizations, transactional leadership behaviors seem to be generally effective within projects. Previous research has indicated a positive relation between transactional leadership factors and followers’ commitment and performance (Antonakis et al., 2003) which is decisive for the success of a project. We therefore propose the following hypothesis:
Transformational Leadership and Temporary Organizations
If transactional leadership emphasizes an exchange and task-oriented leadership, then transformational leadership focuses on the transformation of followers’ values and beliefs to attain the organization’s objectives. With charisma as one of the core elements, transformational leaders create a long-term vision (Christenson & Walker, 2004) and inspire their subordinates to commit (Herold, Fedor, Caldwell, & Liu, 2008) to the organizations’ tasks. Because projects are time-limited undertakings, these leaders must find ways to include long-term aims into short-term projects (Love, Fong, & Iraní, 2005). Transformational leaders emphasize this long-term vision and mission (Herold et al., 2008). Transformational leadership hinges on the followers’ capacity to identify with the leader (Jung & Avolio, 2000) and may induce success even if the coordination mode is unclear (i.e., missing hierarchies) and the team constellations are unstable. Charisma as an important element of transformational leadership seems to be particularly important to projects, as project leaders generally have little de facto authority and cannot use certain methods to influence their followers (e.g., rewards and punishments; Hodgetts, 1968). In addition, a project’s participants may stem from different organizational, educational, and national backgrounds (Parker & Skitmore, 2005). More recently, Gundersen et al. (2012) presented a study on the effects of transformational leadership in international project teams. They found a positive effect of transformational leadership on team performance which was negatively moderated by the dynamism of the work context.
Studies based on various neocharismatic theories have confirmed that transformational leadership leads to high levels of follower motivation and commitment, especially in times of uncertainty (House & Aditya, 1997). In addition, transformational leaders inspire followers to solve problems and find new ways of dealing with challenges (Gumusluoglu & Ilsev, 2009; Rafferty & Griffin, 2004). Because of the unique outcomes that often accompany projects (Brockhoff, 2006), realizing the project goals often calls for novel (Marks et al., 2001) and uncertain (Atkinson, Crawford, & Ward, 2006) practices. As these practices require creativity (Pich et al., 2002) and innovation (Uhl-Bien, Marion, & McKelvey, 2007) from the project’s participants, the leader must encourage these traits and allow participants to exercise some degree of independence (Heinz, Baga, Gebert, & Kearney, 2006; Morgeson, 2005) rather than merely giving instructions (Goodman & Goodman, 1976). Transformational leaders must also play the roles of coach and advisor. As temporary teams often consist of individuals with complementary skills but originating from different departments (Zwikael & Unger-Aviram, 2010), a one-on-one relationship (denoted as individual consideration) enhances the commitment of these individuals to the tasks that must be accomplished (Emery & Barker, 2007). Paying special attention to each individual also fosters two-way exchange and communication. These relationships partially compensate for the absence of hierarchies and the temporary nature of the collaboration (Cleland, 1967). Taking all of the above considerations into account, we propose the following:
The Joint Effects of Transactional and Transformational Leadership
Because the characteristics of temporary organizations result in a high level of complexity, Bass (1999) argues that transformational leadership expands the range of leadership behaviors that must be displayed. At the same time, transformational leadership has been shown to be one of the most effective leadership styles in times of change and uncertainty (Bass, 1990). Therefore, this leadership style seems of particular interest to the temporary organizations as bearers of change for a permanent organization (Sydow et al., 2004). We thus propose the following hypothesis:
Commitment and Project Success
Employees with high levels of affective commitment toward an organization are found to value their membership within the organization because they identify themselves with that organization (McShane & Glinow, 2008). Thus, affective commitment is the strongest predictor of an employee’s effort to support an organization (Meyer & Herscovitch, 2001). Steyrer, Schiffinger and Lang (2008) found that organizational commitment forms a link between leadership and organizational performance. This commitment can be displayed toward various levels and objects within an organization even if these factors are not necessarily related to each other. Meyer, Allen, and Smith (1993) showed that a strong commitment toward a task does not necessarily imply that the employees have a strong commitment toward the overall organization. Nevertheless, affective commitment indicates that an individual desires to support the object of his or her commitment (Meyer & Herscovitch, 2001). Hence, an affective commitment to a particular change positively influences the participant’s willingness to actively participate in the change (Herold, Fedor, & Caldwell, 2007). We may therefore conclude that affective commitment to change is a required characteristic for fulfilling the given tasks. Because projects inherit change, they are established as external supplements to an organization’s routine processes (Lundin & Söderholm, 1994). As project members are vital to the realization of any given tasks within a project, projects’ leaders must influence their followers’ commitment to produce a positive project outcome (Hoegl et al., 2004). We thus propose the following hypothesis:
The Moderating Role of Project Complexity
Temporary organizations are often formed in response to dynamic environments (Ekstedt et al., 1999). Firms may call projects into existence because they realize their goals cannot be effectively reached within the structure of the permanent organization (Sydow et al., 2004). Consequently, we have to consider the contextual factors when we analyze temporary organizations (Gundersen et al., 2012). The uniqueness of a project often calls for the inclusion of novel tasks that are significantly different from an organization’s routine tasks (Brockhoff, 2006). In addition, each participant has expert skills that are needed to fulfill the projects’ tasks. As these experts stem from different organizational units, former relations between the participants often do not exist (Sydow et al., 2004). Both the novelty of the task and the heterogeneity of the participants increase the complexity of the tasks at hand. Neither an established hierarchy nor a long-established sense of coherence exists, both of which could compensate for the absence of routines and the general complexity of the mission (Ratcheva, 2009). Taking all these factors into consideration, the participants may experience feelings of risk and uncertainty (Tatikonda & Rosenthal, 2000). Because temporary organizations typically cause disruptions, the resulting perceptions of change and uncertainty are even stronger in these organizations (Sydow et al., 2004). For this reason, we hypothesize that a project’s complexity and the accompanying feelings of uncertainty play a major role in the effectiveness of leadership in these projects. Prior research has also presented evidence suggesting that an individual’s uncertainty exerts a negative influence on commitment (Hui & Lee, 2000). A transactional leadership style that supervises the day-to-day tasks (Bass et al., 2003) will most likely fail to meet the demands that a project’s complexity may impose on the project’s leaders. Previous research has repeatedly confirmed the effectiveness of transformational leadership in times of uncertainty and change (Bass, 1990). Shin and Zhou (2007) showed that transformational leadership positively moderates the impact of team heterogeneity by improving the relation between the participants’ varying educational backgrounds and the team’s creativity. Studies thathaveanalyzed the positive effects of transformational leadership have generally encompassed not only the nature of the tasks to be accomplished but also the conditions under which these actions took place (Atkinson et al., 2006; Waldman et al., 2001). Accordingly, we assume that project complexity amplifies the effect of transformational leadership within temporary organizations. We therefore suggest the following hypotheses:
Following Liu (1999) and Locke and Latham (1990), we also assume that, as a project’s complexity increases, the project’s transformational leadership exerts a stronger effect on commitment. As an individual’s expectations of his or her performance capability are affected by the project’s difficulty, the individual will adjust his or her behavior (in terms of effort and direction) to ensure the project’s success. Given a high level of commitment, difficult goals are expected to lead to higher performances (Liu, 1999). Thus, we state the following:
Figure 1 integrates Hypotheses 1 through 6 into a research model.
Research model.
Method
Sample and Data Collection
To empirically test our hypotheses, we chose an interindustry study design in order to control for potential industry effects. The challenge in selecting the sample lies in the need to only target employees with a substantial record of project work. We could not use conventional company databases to send out questionnaires as these databases usually do not comprise contact persons with a project experience. Collaboration with the International Project Management Association (IPMA) enabled us to ask respondents with a record in project work. The IPMA member associations in Austria, Switzerland, and Germany allowed us to use their regular newsletter e-mail for distributing our questionnaire. This newsletter informs members about new developments and events related to the associations’ activities. The members of the IPMA member association in Germany (5,300), Austria (600), and Switzerland (550) received the newsletter comprising a link to an online questionnaire. Respondents were asked to report on the last completed project they participated in. A drawback of this approach is that we cannot exactly determine the response rate as we do not know how many newsletter recipients actually read it. Of the 6,450 individuals who received the newsletter 622 followed the link to our questionnaire. Finally, we gathered 163 analyzable data sets from the respondents in Germany (124), Austria (19), and Switzerland (19). The average age of the participants was 42 years, and the participants had 10 years of project experience on average. The average time spent in the current company was 7 years. In total, 78% of the respondents were male. Following the suggestions of Armstrong and Overton (1977) to evaluate the incidence of a potential nonresponse bias, two-sample t tests on dependent variables were performed. Both evaluations indicated that early and late respondents do not differ from one another at the 5% significance level, suggesting that nonresponse bias is not a concern. Table 2 provides an overview of the sample’s composition with respect to the industry structure and the types of projects interrogated by our study.
Sample Composition.
Measures
We mainly used validated scales and items of leadership research based on the Multifactor Leadership Questionnaire (Antonakis et al., 2003; Bass & Avolio, 1990). However, the project management associations supporting our research prompted us to keep the questionnaire as short as possible. Therefore, we could not measure the full range model of leadership (Avolio, 2004) which includes transformational and transactional leadership as active leadership behaviors and laissez faire leadership as passive behavior (45 items in total). We excluded passive leadership as our hypotheses only pertain to active leadership, which is particularly relevant for the context of temporary organizations where new and innovative solutions must be created. For transactional and transformational leadership we selected those items which we believe correspond best to the project context and the hypotheses. Likewise, for measuring project commitment we could resort for measuring commitment to organizational change (Herscovitch & Meyer, 2002) which we adapted to the project context.
Combining two fields of research—leadership studies and temporary organizations/project management—we face varying degrees of methodological maturity. Whereas leadership research pays much attention to the validation of scales, studies on projects and other forms of temporary organizations often attach less value to these methodological questions with no validated scales that represent generally accepted measurement instruments. We, therefore, developed our own measures based on the literature on project complexity and project success.
All items used to measure the constructs are included in the appendix. Likewise, the appendix shows the results of tests for validity and reliability which indicate that all measures we use meet the usual criteria.
Transactional Leadership
The research results are discordant with regard to the factors that constitute transactional leadership behavior (Bycio, Hackett, & Allen, 1995). Several researchers have remarked that the factors constituting transactional leadership, which were originally proposed by Bass (1995), are not reflected in empirical research (Antonakis et al., 2003; Den Hartog et al., 1997). Given the characteristics of each project, we aimed to capture the factors of transactional leadership that have previously been described as active leadership behaviors (Rafferty & Griffin, 2004). We tested for two leadership behaviors, contingent reward and management by exception. Similar to Waldman et al. (2001), we used a 5-point Likert-type scale. To assess each dimension separately, we operationalized transactional leadership as a second-order construct of Type 2, with reflective measures for the first-order factors and formative measures for the second-order factors (Bassellier & Benbasat, 2004; Jarvis, MacKenzie, & Podsakoff, 2003). More specific, the second-order factors were contingent reward and management by exception.
Transformational Leadership
Although most researchers have referred to transformational leadership as comprising four factors, the existing research has shown little agreement on the actual range of factors (Avolio et al., 2009; Ayoko & Callan, 2010). As explained above, transformational leadership behaviors such as charisma (i.e., idealized inspiration), individualized consideration, and intellectual stimulation appear to be most desirable, given the prevailing characteristics in a project’s setting. We decided not to include idealized influence as a fourth dimension of transformational leadership (Avolio, 2004). This dimension mainly relates to the ethical behavior of the leader (personalized versus socialized). Although this dimension is important to see if the leader follows a higher morale, we believe that it is less relevant for a project context which is characterized by high time pressure and a very dynamic environment. We operationalized transformational leadership as a second-order construct of Type 2 with reflective measures for the first-order factors and formative measures for the second-order factors (Jarvis et al., 2003) by using Waldman et al.’s (2001) 5-point Likert-type scale for charisma and Bycio et al.’s (1995) scales for individualized consideration and intellectual stimulation.
Commitment
Because affective commitment best indicates the aforementioned positive effects that an individual’s commitment exerts on a given task or organization (McShane & Glinow, 2008; Meyer et al., 1993), we used Herscovitch and Meyers’s (2002) construct to measure the affective commitment to change with a 5-point Likert-type scale. The wording was adjusted to the context of projects.
Project Success
Although project success is a relevant variable for most researchers, the factors of project success are not clearly defined (Prabhakar, 2008). The absence of a universal definition for the dimensions constituting project success (e.g., Lipovetsky, Tishler, Dvir, & Shenhar, 1997; Dvir, Lipovetsky, Shenhar, & Tishler, 1998; Shenhar, Levy, & Dvir, 1997) makes measuring this concept even more difficult. To fully grasp the general definition of project success (e.g., Ashley, Lurie, & Jaselskis, 1987; Pinto & Slevin, 1988), we used a scale based on Collins and Baccarini’s (2004) comprehensive concept for measuring project success. Accordingly, project success consists of two separate dimensions (Collins & Baccarini, 2004): First, project efficiency refers to the so-called “iron triangle” and measures if the project meets the predetermined constraints in terms of time, cost, and quality. Second, project effectiveness measures if the project meets the objectives of the project set by different stakeholders. We asked respondents to rate the perceived success of the project in terms of the iron triangle, the overall perception of success in terms of efficiency measures, and the perceived satisfaction of the project stakeholders. All these items were measured using a 5-point Likert-type scale.
Project Complexity
In a review of research on project complexity Geraldi, Maylor, and Williams (2011) distinguish between different types of complexity in a project context. We built on these types and measured task complexity with two items “the project had a high degree of task novelty” and “the project had a high degree of complexity concerning content.” Structural complexity resulting from heterogeneity and diversity of project teams was measured with the following item: “To me, the project had a high degree of complexity concerning interdisciplinary participants.” We measured complexity related to risk and uncertainty with one item: “The project was characterized by high risk and uncertainty.” All items were measured on a 5-point Likert-type scale.
Analysis
We tested the cause-and-effect chains by applying structural equation modeling. As the size requirements of the sample are considerably smaller than the minimum size recommended for covariance-based techniques, especially for complex models (Chin & Newsted, 1999), we deployed SmartPLS 2.0 (Temme, Kreis, & Hildebrandt, 2010) to estimate the parameters in the outer and inner model and used a path weighting scheme for the inside approximation (Chin, 1998; Tenenhaus, Vinzi, Chatelin, & Lauro, 2005). To obtain the standard errors of the estimates, we used the nonparametric bootstrapping (Chin, 1998; Efron & Tibshirani, 1986; Tenenhaus et al., 2005) implemented in SmartPLS 2.0 with 500 replications and the construct level changes preprocessing option. We set up the higher order latent variables by implementing a two-stage approach for our estimation process (Henseler, Wilson, Giitz, & Hautvast, 2007). Within the two-stage approach, we initially estimated the latent variable scores without the presence of the second-order construct. Then, we used the latent variable scores as indicators in a separate higher order structural model analysis.
Our study depends on single informants who provided subjective perceptions of the project’s characteristics and rated the perceived behavior of the project leader. This may lead to a common method bias, that is, the potential problem that the observed variance may partly be attributed to the measurement method instead of the constructs (Doty & Glick, 1998). We followed the recommendations of Podsakoff, MacKenzie, Lee, and Podsakoff (2003) and combined procedural with statistical remedies to control for a potential method bias. On the procedural level, we clearly separated dependent and independent variables in the questionnaire. We reduced the ambiguity of each item’s construction and kept the questions as simple as possible (Jarvis et al., 2003). On the statistical level, we applied Harman’s single-factor test (Podsakoff et al., 2003). Therefore, we implemented all independent and dependent variables in an exploratory factor analysis. As no single factor did emerge and the first factor only accounted for 48.55% of the total variance and thus falls below the common threshold of 50% (Podsakoff et al., 2003; Podsakoff & Organ, 1986; Tiwana & Keil, 2007), the test provides additional assurance that our results are not due to common methods variance. Furthermore, we used the Lindell–Whitney marker variable test, implementing a theoretically unrelated marker variable within our model (Lindell & Whitney, 2001). High correlations among any of the study’s principal constructs and the marker variable would indicate common methods bias (Malhotra, Kim, & Patil, 2006). We separately repeated the test with three marker variables theoretically unrelated to our principal constructs: (a) the age of the respondent, (b) environmental uncertainty, and (c) the number of hierarchical levels in the organization. The average path coefficients between the study’s principal constructs and all marker variables, namely age of the respondents (β = 0.054), environmental uncertainty (β = 0.057), and number of hierarchical levels in the organization (β = 0.046), were low and nonsignificant, thereby providing no evidence of common method bias (Tiwana, 2008). Having implemented both procedural and statistical remedies to control for common method bias, we conclude that a common method bias did not significantly affect our results.
Results
The estimations fit the data well. The model explains 25.4% of the variance in the respondents’ commitment to the project and 37.7% of the variance in project success. Because the square root of the average variance extracted (AVE) exceeds the intercorrelations of the construct with the other constructs in the model, the discriminant validity is supported (Chin, 1998; Fornell & Larcker, 1981; Hulland, 1999; for more detailed information, see Table 3). With respect to the variance inflation factors (VIFs), all the formative indicators are far below the rule of thumb provided by Henseler, Ringle, and Sinkovics (2009). The final items included in the survey and their sources appear in the appendix. To support the relevance of our model, we estimated the predictive power by applying a blindfolding approach (Fornell & Bookstein, 1982). Because this approach results in Q2-values of .14 (project commitment) and .23 (project success), both of which are significantly different from 0, we may conclude that the model has predictive power (Geisser, 1974; Stone, 1974). All the indicator reliabilities are greater than .65, the composite reliability is greater than .86 and the AVE is greater than 0.52 in all cases (see Table 3). The calculated variance inflation factors for transformational leadership, transactional leadership, and project commitment are less than 1.5. Thus, no multicollinearity is present at the structural level of the model.
Descriptive Statistics and Measurement Model Results at Structural Model Level.
Note. SD = standard deviation; CR = composite reliability; AVE = average variance extracted.
Significant at p < .01.
The empirical results support all hypotheses about the inner model relationships (see Figure 2), except for Hypotheses 5a and 6. Specifically, the results show that both transformational (β = 0.31, p < .01) and transactional leadership (β = 0.15, p < .1) are positively related to project commitment. This finding supports Hypotheses 1 and 2. Consistent with Hypothesis 4, we found a significant, positive effect of project commitment on project success (β = 0.58, p < .01). Although not explicitly stated within our hypotheses, we also controlled for direct effects of transformational and transactional leadership on project success. However, no direct effect of either transformational (β = 0.12, ns) or transactional leadership (β = −0.06, ns) was present. As the direct effects might have been mediated by project commitment, we implemented a further mediation analyses to test the mediating role of project commitment. We thereby followed the procedures suggested by Iacobucci and Duhachek (2004), who built on Baron and Kenny’s (1986) recommendations for testing mediation effects (Helm, Eggert, & Garnefeld, 2010). As the direct effects on project success of both transformational and transactional leadership turned out to be nonsignificant when modeled along with the selected mediator variable project commitment, our results indicate the establishment of a complete mediation (Baron & Kenny, 1986). To test each proposed mediation effect for significance, we additionally calculated the z statistic (Sobel, 1982). The results of two Sobel tests show that project commitment significantly mediates the effect of both transformational (z value = 2.45, p < .01), and transactional leadership (z value = 2.10, p < .05) on project success.
Schematic representation of empirical results.
The inclusion of the interaction term, project complexity, positively affected the influence of both transformational (β = 0.16, p < .1) and transactional (β = 0.19, p < .05) leadership on project commitment. These findings do not confirm Hypothesis 5a but do support Hypothesis 5b. The results indicate that project complexity exerted a negative impact on the relationship between transformational (β = −0.15, p < 0.05) and transactional (β = −0.13, p < .1) leadership. We also found that transformational and transactional leadership indirectly affected project success, with transformational leadership having a stronger effect (β = 0.18) than transactional leadership (β = 0.084). However, this result lies outside the scope of our hypotheses. Table 4 reports the effects of the covariates’ project involvement, internal versus external project, project size, and project role which we used as control variables. Figure 2 provides an overview on the empirical results.
Covariates and Significance Level.
Although we confirmed that “Internal versus external project” (β = −0.13, p < .05) and “Number of external members (EM)in project” (β = −0.161, ns) exert weak negative effects on project commitment, no effects for “Degree of project involvement” (β = −0.09, ns), “Project size” (β = −0.05, ns), “Project duration” (β = −0.00), and “Mode of involvement” (β = -0.161, ns) were found. If one of the exogenous latent variables in the structural model is either excluded or included, then the relative changes in the goodness-of-fit R2 define the effect size. According to Chin (1998), the ƒ2 values of 0.02, 0.15, and 0.35 specify whether the latent exogenous variables have a small, medium, or large effect, respectively. The size of the effect that transformational and transactional leadership exert on project commitment was confirmed to be low (f2 Transformational = 0.049; f2 Transactional = 0.013), but the effect size of transformational leadership is more than three times larger than that of transactional leadership. Thus, Hypothesis 3 is supported.
Discussion
We presented the result of a study on leadership in temporary organizations. As temporary organizations bring change to an overall organization (Sydow et al., 2004), the project’s participants are exposed to a large degree of change and uncertainty (Gällstedt, 2003), which in turn imposes a high probability of low follower commitment (Armenakis, Harris, & Mossholder, 1993). We examined the effects of transactional and transformational leadership behavior to investigate the impact of leadership on follower commitment in this organizational context. We found that task and people-oriented leadership behaviors are important in project settings. Consistent with our hypotheses, we showed that transactional and transformational leadership styles exert positive effects on a follower’s commitment to a project. Although transactional leadership and transformational leadership alone proved to only exert minor effects on commitment, the full range of active leadership behavior (Avolio, 2004; Bass, 1990) explained a fourth of the variance in commitment.
The present study makes three major contributions to the literature. First, we extended the previous findings on the applicability of transactional and transformational leadership in different organizational contexts by examining these leadership styles in the context of temporary organizations (i.e., projects). Second, we analyzed in detail the characteristics of these settings and their effects on project leadership. Third, we tested the hypothesized causalities on the basis of different project and industry types. Our findings carved a path for future studies within these particular contexts.
The Role of Transactional Leadership in Projects
Previous scholars have claimed that transactional leadership (i.e., task-oriented leadership) serves as the basis of all leadership in organizations (Bass, 1990). This declaration is easy to understand, as the primary objective of organizational leadership is to achieve both high efficiency in the use of resources and high consistency in organizational operations, products, and services (Yukl, Gordon, & Taber, 2002). Following Bass (1990), we regarded transactional leadership as the basis for the display of transformational leadership. Our analysis confirmed that transactional leadership positively correlates with transformational leadership and that the latter can realize its full effectiveness through this relationship. Consistent with our hypotheses, we found that transactional leadership also directly affects commitment. This finding supports the assumption that a structured course of action exerts a positive effect on follower commitment. This structure reduces uncertainty and compensates for missing routines by providing clear assignments. These results contribute to the existing research on leadership in temporary environments by clarifying the general role of transactional leadership in temporary organizations. Prior studies have either regarded transactional leadership as relevant to single project types (e.g., Müller & Turner, 2010) or focused on the role of transformational leadership (e.g., Keegan & Den Hartog, 2004; E. Wang, Chou, & Jiang, 2005). Therefore, our study complements the findings of prior research by elucidating the role of transactional leadership in projects: Neither transactional leadership (i.e., task-oriented leadership) nor transformational leadership (i.e., person-oriented leadership) is more effective in projects. Transactional leadership serves as the foundation for leaders seeking to maximize their influence on their followers’ commitment to a project. In contrast to our hypothesis, which stated that transactional leadership experiences decreased importance in projects with high complexity, we found that transactional leadership positively affects commitment. This finding underlines the general importance of transactional leadership.
The Role of Transformational Leadership in Projects
As predicted, transformational leadership had a strong positive effect on subordinate commitment. In addition, the f2 values show that transformational leadership had a stronger positive effect on commitment than transactional leadership. The results confirm our assumption that transformational leadership is needed to cope with the drawbacks (e.g., missing routines and lack of authority) that accompany projects. Moreover, we found that transformational leadership exerted an indirect positive effect on the perceived success of a project. This finding highlights the importance of transformational leadership to the successful completion of a project. In contrast to the findings of Keegan and Den Hartog (2004), we found a significant link between transformational leadership and commitment to projects. These findings also support the previous research on industry and single project type (Gundersen et al., 2012; Keller, 1992; E. Wang et al., 2005). Our findings confirm the positive effects of transformational leaders on project team attributes and outcomes across both different industries and different project types. This result has important implications for project management practices, as the positive effects of transformational leadership can now be assumed to exist in all projects. In the context of leadership research, these findings shed light on the emergence of transformational leadership. The effectiveness of such leadership behaviors is not only restricted to permanent organizations but can also be important in temporary environments.
The Role of Commitment in Temporary Organizations
Our study supported the relevance of commitment to the overall success of a temporary organization. As prior studies have frequently stated that affective commitment plays a major role in a range of organizational outcomes (e.g., involvement in the organization and task fulfillment), our findings indicate that the same can be said of temporary organizations. In contrast to permanent organizational settings, committing to a temporary project is assumed to be more difficult. This assumption is particularly important because scholars have long assumed that project success is realized through effective administration (Carpenter, 2002). Our findings indicate that, unlike the straightforward management of projects, articulating the higher goals of a temporary undertaking exerts a positive effect on the project’s success, although it does so indirectly by establishing higher levels of commitment.
With regard to managerial implications, these findings complement the knowledge on the importance of effective administration to projects. Although conventional leadership solemnly focuses on the fulfillment of a project’s goals, transformational leadership focuses on the intentions of the project’s participants to fulfill these project goals. Both leadership foci are essential, as a project’s characteristics and complexity could hinder a leader from having the information necessary to giving the right directives. In these cases, committed subordinates, unlike uncommitted employees, would try to do the right thing. However, committed individuals might at some point fail to recognize what needs to be done because they lack information. Thus, both leadership behaviors seem vital to project managers.
The Role of Perceived Complexity in Temporary Organizations
The degree of a project’s complexity reflects a range of project characteristics that can distinguish themselves within the range of projects that we have scrutinized in the present study. With regard to the characteristics of temporary organizations (cf. Table 1), the limited duration of these organizations applies to all kinds of projects and programs but may differ across the different cases. Short-term projects, such as the update of an enterprise-resource planning system, are certainly different from voluminous construction projects. The same is true for the remaining characteristics of temporary organizations. We regarded complexity as amplifying the mode of actions in the projects of our study. We confirmed these assumptions to a great extent, as the results highlight the compelling role of transformational leadership in difficult circumstances. Transformational leadership exerts an increasingly positive effect on commitment in situations of increasing complexity, and the same is true for transactional leadership. At the same time, we found that project complexity negatively affects project success. This finding highlights the importance of both transactional and transformational leadership in countering this effect. Our work also expands on the findings of Müller and Turner (2010), who found a positive relation between leadership styles that encourage engagement and the success of projects with a high degree of technical complexity. Our findings show that complexity in general calls for transformational leadership and, thus, clarifies the mixed outcomes of earlier studies (e.g., Keegan & Den Hartog, 2004; Müller & Turner, 2010; Prabhakar, 2008) on an industry- and project-type-spanning level.
Limitations and Future Avenues of Research
Our study features some limitations that we would like to see addressed in future research. In particular, the relations of leadership, commitment, and project characteristics could be further investigated to acquire a more objective measure of project complexity and to further support the relation between commitment and overall project success. For example, future studies may use nested data to support their findings. Furthermore, as we were somewhat restricted in the use of complete scales measuring transformational and transactional leadership, future research might validate our results using the full measurement inventories. Nevertheless, as our main concern was to analyze the effects of an individual’s perception of leadership behavior on the individual’s commitment, we believe that this research design best analyzed the assumed coherences. We computed the direct effect of project complexity on both leadership constructs while using zero effect size. In doing so, we ruled out any possible cognitive effects related to the lower ability to perceive transformational leadership in projects with high complexity (e.g., Pillai & Meindl, 1998).
Because we modeled project complexity as one factor, we would welcome further research that scrutinizes the effectiveness of transactional and transformational leadership with regard to different types of projects. Additionally, as we tested our hypotheses using projects, other forms of temporary organizations, such as programs, might display divergent mode of actions. Although these organizations basically share the same characteristics, we propose that our findings should be validated by studies of programs, as cooperation in these settings might be different from the causalities tested in this article. For example, repeated collaboration may exist, which could significantly alter the relations between leaders and followers. This article may also lead to a longitudinal study analyzing the changes in the relative importance of transactional and transformational leadership styles over time.
Finally, within our study the same individual answers both the questions to capture endogenous and exogenous variables, which could lead to common method variance as a potential area of concern. For our research design, we implemented both procedural and statistical remedies to minimize possible problems related to common method variance and to control for common method bias. Although no sign for common method bias was detected, future research might use a dyadic design in which the individual who answers the dependent variables is different from the individual who answers the dependent variable. Another option is to implement research designs in which project success is assessed with a different instrument than the independent variables, for example, using secondary data on financial project outcomes to capture project success.
Conclusion
Leadership research has called for more empirical research on leadership in different organizational contexts (Antonakis et al., 2003; Avolio et al., 2009; Porter & MacLaughlin, 2006). Following this call, we presented the results of a study on the effects of leadership in temporary organizations. Not only are projects and programs as forms of temporary organizing increasing in their frequency of occurrence, but permanent organizations are also increasingly adopting temporary elements for adapting to dynamic environments. Temporary organizations often accompany change in the organizational structure of a permanent organization and they are used to generate innovations (Whittington, Pettigrew, Peck, Fenton, & Conyon, 1999). The present study highlights the importance of transactional and transformational leadership in times of change, complexity, and uncertainty. Although this concept has been investigated within the context of permanent organizations (e.g., Bordia, Hobman, Jones, Gallois, & Callan, 2004) and with respect to the role of higher management (e.g., Waldman et al., 2001), we showed that these leadership approaches also apply to the context of temporary organizations. As we scrutinized the assumed coherences on a broad basis of industry and project types, our results are not only generalizable but also serve as an extension of the studies that have been conducted in single industries (e.g., Gundersen et al., 2012; Thite, 2000) and project types (e.g., Tatikonda & Rosenthal, 2000) as well as those that delivered mixed results (e.g., Keegan & Den Hartog, 2004). Leaders in temporary settings may use transformational leadership to influence their subordinates’ commitment to the project. Nonetheless, they must be aware of the full range of leadership behaviors as they engage in increasingly complex situations.
Footnotes
Appendix
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.