Shared Leadership and Team Effectiveness: An Investigation of Whether and When in Engineering Design Teams

This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

Associated Data

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

Abstract

Shared leadership is lauded to be a performance-enhancing approach with applications in many management domains. It is conceptualized as a dynamic team process as it evolves over time. However, it is surprising to find that there are no studies that have examined its temporally relevant boundary conditions for the effectiveness of the team. Contributing to an advanced understanding of the mechanism of shared leadership in engineering design teams, this research aims to investigate whether shared leadership is positively related to team effectiveness and when shared leadership is more likely to be effective. Using a field sample of 119 individuals in 26 engineering design teams from China and the technique of social network analysis, we found that, consistent with cognate studies, shared leadership is positively related to team effectiveness when measured in terms of team task performance and team viability. Moreover, by integrating the project life cycle as a moderator, this study is among the first to investigate the temporal factors, for the effectiveness of shared leadership. The result shows that the stage of the project life cycle moderates the positive shared leadership-team effectiveness relationship, such that this association is stronger at the early phase than at the later phase of the project. Overall, these findings offer insightful thoughts to scholars in the field of shared leadership and bring practical suggestions for project managers in business who seek to implement best practice in organizations toward high team effectiveness.

Keywords: shared leadership, team effectiveness, project life cycle, social network analysis, engineering design teams

Introduction

In recent years, leadership researchers have emphasized a team-level phenomenon, where leadership is carried out by the team as a whole, rather than exclusively by those at the top or by those in formal leadership positions (Carson et al., 2007; Pearce et al., 2014). As such, the notion of shared leadership has gained more traction in the extant literature. By definition, shared leadership is described as “a dynamic, interactive influence process among individuals in groups for which the objective is to lead one another to the achievement of group or organizational goals or both” (Pearce and Conger, 2003, p. 1). As Acar (2010) noted, shared leadership represents a fundamental shift away from the notion of a single, appointed leader, to the idea that team members mutually influence each other and collectively share leadership roles, responsibilities and functions. Recent empirical work has provided evidence for the important role of shared leadership in groups (Nielsen and Daniels, 2012; Nicolaides et al., 2014; Sousa and Van Dierendonck, 2016; Sun et al., 2016). More interestingly, some studies have even found that shared leadership is more influential than convectional vertical leadership for team effectiveness (Pearce and Sims, 2002; Ensley et al., 2006). However, our understanding of whether shared leadership is positively related to team effectiveness and when shared leadership is more likely to be effective is still limited in at least three fundamental ways.

First, in recent years, researchers and practitioners have advocated the benefits of shared leadership as a way to promote team effectiveness. For example, Ramthun and Matkin (2012) stated that shared leadership is often advantageous, since members are more likely to follow the person having the best knowledge and skills than depending solely on the vertical influence process of traditional leadership. Indeed, many other empirical studies have also demonstrated that teams with shared leadership yield higher team effectiveness (Pearce and Sims, 2002; Wang et al., 2014; Serban and Roberts, 2016). However, we must caution that this is not always the case. Fausing et al. (2013) and Mehra et al. (2006) failed to find support for this significant and positive relationship, and Boies et al. (2011) even found that shared leadership exerts a negative influence on team effectiveness. Such inconsistent findings point to the need for more empirical evidence. Therefore, in order to enrich our understanding of the value of shared leadership, the first purpose of our study is to explicitly examine the shared leadership – team effectiveness relationship. In this study, we define team effectiveness as the extent to which teams meet the expectations of organizations (Essens et al., 2009). This viewpoint encourages us to think about team effectiveness from a multidimensional perspective. Consequently, we follow Aube and Rousseau (2005), Balkundi and Harrison (2006), and Mathieu et al. (2008), who consider team effectiveness from two distinct aspects: team task performance and team viability. Team task performance refers to how well the group meets (or even exceeds) work expectations while team viability is the potential of teams to retain its members and to function effectively over time (Balkundi and Harrison, 2006).

Second, in order to gain a more fine-grained understanding of the impacts of shared leadership, unanswered questions must be addressed. More specifically, there is a clear need to investigate the temporally relevant moderators for its effectiveness. Researchers have emphasized that shared leadership is a dynamic, emergent, time-varying construct (Avolio et al., 2009) that is affected by the environment of a team (Carson et al., 2007; Wu et al., 2020) and task characteristics (Serban and Roberts, 2016; Hans and Gupta, 2018). Therefore, continuous changes in the inputs, processes and outputs of different phases of the project life cycle could influence the emergence of shared leadership in teams (Wu and Cormican, 2016) as well as its relationship with team effectiveness. However, the potential moderating impact of the project life cycle for the effectiveness of shared leadership is not well theoretically developed nor rigorously empirically tested. This important unaddressed gap needs further attention so as to provide insights into the boundary conditions regarding when shared leadership is more or less influential to team effectiveness. Consequently, the second research goal is to focus on the dynamic nature of shared leadership and investigate the moderating effect of the project life cycle in the relationship between shared leadership and team effectiveness.

Third, although there is growing interest in the shared leadership domain, studies concentrating on project teams are still limited and under-developed (Scott-Young et al., 2019). Shared leadership theory has been widely spread and applied across a range of team types, e.g., top management teams (Singh et al., 2019), entrepreneurial teams (Zhou, 2016), consulting teams (Carson et al., 2007), and change management teams (Pearce and Sims, 2002). However, there is a dearth of investigations relating to project teams. While the current workplace is becoming increasingly project-centric (Scott-Young et al., 2019), there remain very few studies focusing on shared leadership theory in the project management context. In order to extend the external validity of the shared leadership construct in project settings, this study examines the effectiveness of shared leadership in project-based engineering design teams. Moreover, as project teams uniquely have definitive start and end times based on the duration of the tasks (Farh et al., 2010), it is well suited to help explain when shared leadership is more likely to be effective in teams.

Taken together, this research seeks to enrich our understanding of the mechanisms of shared leadership and investigates whether and when shared leadership is positively related to team effectiveness in engineering design teams. To do this, we used the social network approach to measure the construct of shared leadership by calculating network density and creating binary matrices as well as sociograms. Team effectiveness was measured using nine items consisting of two separate, theoretically derived subscales: team task performance and team viability. Moreover, an internal consistency analysis and confirmatory factor analysis was performed to assess the reliability and validity of our measurement model. We then conducted a two-way moderated hierarchical regression analysis (Carson et al., 2007; Erkutlu, 2012; Fausing et al., 2013) in this study so as to test hypotheses proposed. By doing so, our study makes several significant contributions: (1) it extends a line of research and explicitly examines the relationship between shared leadership and team effectiveness; (2) it builds on the dynamic nature of shared leadership and is among the first to investigate an important temporal moderator, the project life cycle, for the effectiveness of shared leadership; (3) it adds to the academic debate by extending the external validity of shared leadership theory in engineering design teams; (4) it brings insightful thoughts to the field of project management by providing practical suggestions for project managers in business who seek to implement best practice in their organizations.

Theory and Hypotheses

Shared Leadership Theory

Leadership scholars have realized the importance of shared leadership and worked to understand how to conceptualize it, measure it, and to assess what impacts it brings to teams. Table 1 presents details of relevant prior empirical studies. As illustrated, conceptually, shared leadership is a team-centric phenomenon (Ensley et al., 2006; Serban and Roberts, 2016) whereby team members engage in “leadership roles and responsibilities on behalf of the team” (Robert and You, 2018, p. 503), and “accepts their colleagues’ leadership” (Aubé et al., 2017, p. 199). Furthermore, shared leadership is not a static process; it is defined as an emergent, dynamic phenomenon that unfolds over time (Avolio et al., 2009; Drescher et al., 2014; Wang et al., 2014). According to Carson et al. (2007), shared leadership is considered in terms of a continuum ranging from low to high, which implies that shared leadership is not a rigid either-or category, but occurs in every group at various levels (Liu et al., 2014).

TABLE 1

Definitions, measures, and impacts of shared leadership.

While progress has been made relating to the definitions of shared leadership, many empirical studies have centered on what impacts shared leadership brings. As shown in Table 1 , the positive relationship between shared leadership and team performance has received much attention (Sivasubramaniam et al., 2002; Ensley et al., 2006; Mehra et al., 2006; Carson et al., 2007; Drescher et al., 2014). Additionally, shared leadership is also demonstrated to be positively related to team functioning (Bergman et al., 2012), team proactive behavior (Erkutlu, 2012), team and individual learning (Liu et al., 2014), team member’ diversity and emotional conflict (Acar, 2010), team members’ trust, autonomy and satisfaction (Robert and You, 2018). These findings are encouraging and suggest the need for more sophisticated designs on the notion of shared leadership. Accordingly, this study extends a line of research to further examine its relationship with team effectiveness and goes beyond simple relationships to investigate when shared leadership plays a stronger or weaker role in the effectiveness of teams. The relevant research hypotheses are proposed below.

Shared Leadership and Team Effectiveness

Based on the work of Aube and Rousseau (2005), Balkundi and Harrison (2006), and Mathieu et al. (2008), team effectiveness is considered in terms of two distinct aspects: team task performance (how well the group meets (or even exceeds) work expectations) and team viability (the potential of teams to retain its members and to function effectively over time). This assessment conforms to the classic work of Barrick et al. (1998), who suggested that a comprehensive assessment of team effectiveness should capture both current team effectiveness (i.e., present task performance) and future team effectiveness (i.e., capability to continue working together). Therefore, this research adopts a broad perspective to team effectiveness and explores the relationship between shared leadership and team effectiveness.

First of all, this study expects that shared leadership is positively associated with team task performance. As suggested by Day et al. (2004), shared leadership advances the social capital of the team via the utilization of team resources such as the knowledge and capability of group members, which subsequently fosters team task performance. Katz and Kahn (1978) also proposed that when group members offer leadership to others and to the mission or purpose of their group, they bring more personal and organizational resources to the task, share more information, and they experience greater commitment. Further, when group members are influenced by their fellows, team functioning is improved as high levels of respect and trust are evidenced among group members. Collectively, teams exhibiting these characteristics, can also exhibit greater levels of performance (Day et al., 2004). This premise aligns with many empirical studies (see Table 1 ). For instance, Carson et al. (2007), in a study of 59 consulting teams, found that shared leadership is positively associated with team performance as rated by clients. Ensley et al. (2006), in a study of 66 top management teams, demonstrated that shared leadership is a more significant predictor than vertical leadership of new venture performance when considered in terms of revenue and employee growth. Furthermore, Drescher et al. (2014), in a longitudinal examination of 142 teams who engaged in a strategic simulation game, also demonstrated support for the positive influence of shared leadership on team task performance. Taken these together, this study proposes:

Hypothesis 1a: Shared leadership is positively related to team task performance in engineering design teams.

Shared leadership, as an important intangible resource available to teams (Carson et al., 2007), fosters not only team task performance, but also team viability. As Wood and Fields (2007) suggested, shared leadership exerts a series of positive impacts on team members’ job perceptions: it brings low levels of role overload, role conflict, role ambiguity and job stress, as well as high levels of job satisfaction. Similarly, Bergman et al. (2012) also demonstrated that teams with shared leadership experience less conflict, greater consensus, and higher intragroup trust and cohesion. This may foster team viability as members in shared leadership teams experience increased interdependence, more collaboration, and they sense greater levels of satisfaction. Additionally, when there is effective coordination and collaboration among team members fulfilling leadership responsibilities, it is easier for them to identify the potential causes of conflicts and propose potential solutions. It thus reduces the amount of conflict and promotes team consensus and trust (Balkundi and Harrison, 2006). As a consequence, team viability, which retains members and maintains good team functioning over time, could be enhanced. This research therefore posits:

Hypothesis 1b: Shared leadership is positively related to team viability in engineering design teams.

Taken these two hypotheses (hypothesis 1a and 1b) together, this study expects that shared leadership will foster team effectiveness by enhancing team task performance and team viability. As Wang et al. (2014) suggested, shared leadership nurtures a collective identity among members of the team and strengthens the level of engagement with and commitment to the group, which in turn enhances team effectiveness. Moreover, Mathieu et al. (2015) mentioned that shared leadership fosters social inclusion and enhances team cohesion, which can, subsequently, facilitate team effectiveness. In light of this, this research suggests:

Hypothesis 1c: Shared leadership is positively related to team effectiveness in engineering design teams.

The Moderating Role of the Project Life Cycle

Notwithstanding the fact that research on the relationship between shared leadership and team effectiveness brings valuable insights into the understanding of shared leadership in teams, there is an important omission in prior studies regarding its temporal moderating roles on such a relationship (Carson et al., 2007; D’Innocenzo et al., 2014; Wang et al., 2014). In an attempt to open the black box, this study seeks to examine a potential moderator of shared leadership, namely the project life cycle, and expects that the positive association between shared leadership and team effectiveness will be stronger at the early phase than the later phase of the project. This is because the focal concern of the early stage is toward planning and strategy generation (Chang et al., 2003; Farh et al., 2010), where project team members are more willing to engage in mutual leadership as they become proactively involved in constructive communication and decision-making (Wu and Cormican, 2016). It thus allows individuals to bring more resources to the task, share more information, and to experience higher levels of commitment (Bergman et al., 2012). Collectively, these consequences would result in greater team effectiveness (Day et al., 2004; D’Innocenzo et al., 2014). Furthermore, as time and resources are less constrained at the early stage (Farh et al., 2010), members are able to take initiative to develop their own leadership abilities as well as to facilitate the leadership skills of others, which subsequently fosters the effectiveness of project teams (Ensley et al., 2006; Serban and Roberts, 2016). However, when the project advances into the later stage, resources are dedicated to execute project plans (Farh et al., 2010). This leads to a change in the leadership distribution from many team members to a few individuals, who assume the responsibility of integrating resources, controlling the development of the project to meet deadlines and keeping costs within budget (Wu and Cormican, 2016). Teams may no longer afford to spend too much time cultivating a positive team environment to promote shared leadership (Carson et al., 2007). As such, any potential of shared leadership for enhancing team effectiveness would be more difficult to realize in the later stage of the project life cycle. Therefore, this research expects that:

Hypothesis 2: The stage of the project life cycle moderates the positive association between shared leadership and team effectiveness, such that this relationship will be stronger at the early phase than at the later phase of the project in engineering design teams.

Methodology

Research Setting and Sample

A survey-based design was conducted in this study. The sample comprised 26 project-based engineering design teams working in the construction industry in China. As suggested by Carson et al. (2007), shared leadership is effective for teams composed of knowledge-based employees, because people having high levels of expertise and skills seek autonomy in how they apply their specialties, and thus desire more opportunities to shape and participate in the leadership functions for their groups. Engineering design teams comprising knowledge workers have the potential to leverage the expertise of a diverse group of members by pooling their talent and knowledge. This kind of team is likely to nourish the emergence or development of shared leadership. This perspective thus adds to the academic debate on the relationship between shared leadership and team effectiveness and extends the external validity of shared leadership theory into engineering design teams. Moreover, we chose a Chinese sample due to the fact that the conceptualization and operationalization of shared leadership is predominantly developed in the Western countries (see Table 1 ) and it remains uncertain whether its theoretical models hold up in Chinese cultural settings. Furthermore, scholars, like Whetten (2009), have called for more attention to be paid to explaining cultural context effects. Therefore, to plug this gap, this study seeks to extend the validity of the shared leadership construct to a Chinese context, whereby its organizational culture differs from Western countries. Specifically, according to Hofstede et al. (2005), the power distance and collectivism in China are rated stronger than in Western cultures. Initially, a pilot test was conducted with 16 employees from three engineering design teams. Based on feedback provided, minor modifications to the survey items were made. Next, 146 members from 34 engineering design teams were invited to participate in this study. Of the 146 participants who received the questionnaire, 127 returned it, yielding an 87% response rate. Teams with less than three members were eliminated from the sample. It resulted in a sample of 119 employees working in 26 project teams. The average team size of the sample is 5.26. The specific participant demographics are outlined in the Table 2 .

TABLE 2

Measures

Shared Leadership

This research study adopted a social network approach to assess the nature of shared leadership. The social network technique is an intrinsically relational method that advocates a natural theoretical and analytical method to modeling the patterns of the relationships among interconnected individuals (D’Innocenzo et al., 2014). This study used the most common index of social network analysis, network density, to explicitly measure the extent to which team members are perceived to be involved in the sharing of leadership (Wang et al., 2014). This popular measurement was employed in many empirical studies of shared leadership (Carson et al., 2007; Lee et al., 2015; Chiu et al., 2016; Serban and Roberts, 2016). Following Carson et al. (2007), this study assessed the level of shared leadership by requiring every team member to rate each of his/her peers on the following question: “To what degree does your team rely on a particular individual for leadership?” A five-point Likert scale was used to measure the level of perceived leadership, where 1, represents “not at all,” and 5, “to a very great extent.” Network density was then calculated by summing all of the responses from group members divided by the total number of possible relations among group members (Carson et al., 2007; Mathieu et al., 2015). The values of density ranged from 0 to 1, where higher values indicate higher degrees of shared leadership within a team. Furthermore, as shared leadership is a team-level phenomenon, agreement among the respondents’ ratings of group members was also measured thus proving appropriate interrater reliability [mean r_wg = 0.75, ICC(1) = 0.44, ICC(2) = 0.77].

To visually represent the density of shared leadership, this study developed leadership sociograms for each sample team similar to Carson et al. (2007) and Pastor and Mayo (2002). To do this, binary matrices were created, which were then used to quantify the degree of leadership influence for each team and to represent the presence or absence of leadership relations between pairs of team members. More specifically, the raw leadership ratings collected from each participant were aggregated and included in g × g squared matrices. These data were then dichotomized, where values of 4 (to a great extent) or 5 (to a very great extent) are considered as 1, and values of 3 and less are given a value of 0. The second step was to create leadership sociograms based on these binary matrices. Figure 1 shows the leadership sociograms in our study. Specifically, it illustrates three examples with low, middle and high levels of density of shared leadership networks. Among all of our sample data (26 engineering design teams), 0.52 is the lowest score, 0.66 is the medium score, and 0.75 is the highest score of network density. The nodes symbolize team members and the arrows represent leadership relations. One arrow points from team member (A) to member (B), indicating that B is perceived as a source of leadership by A. In this vein, two-headed arrows imply that two members perceive each other as a source of leadership.

Leadership sociograms in this study. Low degree of shared leadership Medium degree of shared leadership (Network density = 0.52) (Network density = 0.66). High degree of shared leadership (Network density = 0.75).

Team Effectiveness

Team effectiveness was measured by team participants (including team leaders and members) via nine items consisting of two separate, theoretically derived subscales: team task performance and team viability using a five point Likert scale ranging from 1 “strongly disagree” to 5 “strongly agree.” Team task performance was assessed using five items derived from Sousa and Van Dierendonck (2016) and Suprapto et al. (2018). It measures the degree to which the project meets its goals, quality, schedule, budget, and overall level of customer satisfaction. Team viability was measured using four items derived from Aube and Rousseau (2005). These include the extent of a team’s capacity to solve problems, the ability to integrate new members, the ability to adapt to changes, as well as the ability to continue to work together in the future. In order to test for the discriminant validity, a confirmatory factor analysis (CFA) was performed. This yielded a good fit to the data (X 2 ₂₇ = 33.90, CFI = 0.99, GCI = 0.94, AGFI = 0.09, RMSEA = 0.05). These CFA results demonstrate the support for the hypothesized structure to measure team effectiveness. This study further examined the correlation between these two subscales to check the convergent validity of this measurement model. The finding provides evidence that these two subscales are highly correlated with each other (r = 0.92, p < 0.001). Given the strong support of the hypothesized measurement model, this study aggregated these two subscales to the group level and then averaged the scores to generate a single variable to represent team effectiveness (Cronbach α = 0.95). To justify whether this aggregation is appropriate, this research used the interrater agreement statistic, r_wg (James et al., 1993). The mean r_wg value of 0.82 was much larger than the conventional cut-off value of 0.70 (James et al., 1993), which implies that on average, there is a high degree of agreement among different raters with a group. Furthermore, the intraclass correlation coefficient, ICC (1) and the reliability of the group-level mean, ICC (2) were also calculated to test between-group variance and within-group agreement (Bliese, 2000). The results showed that the ICC (1) value of 0.73 suggested that team membership accounted for significant variance and the ICC (2) value of 0.92 demonstrated that the group-level means were reliable.

Project Life Cycle

Led by the research of Farh et al. (2010), the phase of the project life cycle was measured from the percentage of the project work completed at the time of the survey, as reported by project managers. In the sample of our study, the mean project completion rate across 26 teams was 56%. This research checked journal guidelines and similar papers (see Farh et al., 2010) and used a mean split, where teams with a percentage of project completion equal to and below 56% were classified as being at an early phase and teams above 56% were classified as being at a later phase. Accordingly, there are 14 project teams in the early phase subgroup with the percentage of project completion ranging from 5% to 56%, and 12 in the later phase subgroup with 57–100% project completion. Figure 2 graphically illustrates the distribution of network density of shared leadership in the early phase vs. later phase.