1 Introduction

Value-based leadership is widely recognized as being very effective for organizational performance across cultures (Meng et al. 2003). It is defined as leadership styles based on strong ideological values (e.g., morality, benevolence) espoused by a leader (Garg and Krishnan 2003; Karakas and Sarigollu 2013). Value-based leadership theory is the major theoretical underpinning that frames this study. Value-based leadership theory describes a leader’s ability to inspire, stimulate, and motivate his/her followers based on core benevolence, morals, integrity, and ethics (Prasad 2016). Previous literature has indicated that a key component of value-based leadership theory is the motivational base of human behavior (McClelland 1985), suggesting leader effectiveness as a function of a specific combination of motives (Meng et al. 2003). Value-based leadership is a value-driven and inside-out undertaking (Hester 2019; Poovan et al. 2006). If a leader emphasizes specific values that appeal to team members, then the members are highly motivated (Meng et al. 2003). Value-based leadership theory focuses on internalized motivation in workers’ ideology, and the influence of such motivation caused by value-based leaders on performance is much larger than the influence of other types of leadership (Jun et al. 2008).

Value-based leaders convey a vital role in the practice of influencing people’s motivation to work together as a team (Niculescu 2014). Specifically, value-based leadership substantially influences two broad categories of motives of employees: dispositional and interpersonal (e.g., Bosselut et al. 2018; Mody and Mody 2012; Raes et al. 2013). Developments in these motives in team dynamics have led to close attention being paid to what is coined as learning goal orientation (i.e., dispositional motive) and interactional justice (i.e., interpersonal motive). Learning goal orientation denotes a team’s dispositional tendency to develop job ability and new skills, whereas interactional justice is the interpersonal treatment received from team leaders with an emphasis on informational and social sensitivity (Heffernan and Dundon 2016). These two factors are simultaneously examined in this study, because previous research has found that the interpersonal context needs to be taken into account to understand what happens in people’s learning when they are on a team (Barron 2003; Van den Bossche et al. 2006).

Although learning goal orientation or interactional justice has been discussed separately in the literature, few studies have taken into account these two variables simultaneously. As an example, Li and Bagger (2008) examined learning goal orientation and justice as two moderators to explain the relationship between role ambiguity and self-efficacy. As another example, Yanghua (2008) presented learning motivation and justice climate as predictors for team learning outcomes. Nevertheless, how learning goal orientation and interactional justice are simultaneously motivated by value-based leadership to influence team performance has not been explored yet, leading to the first research gap that this study aims to fill.

The second research gap to be filled by this study relates to a moderating mechanism of emotional regulation in teams. Emotional regulation represents team workers’ ability to self-manage or self-control feelings and affective attitude, which influence their receptiveness to, and implementation of, surrounding incidents (Kanfer and Heggestad 1997). Emotional regulation is proposed as a moderator in this study, because emotional regulation is a key personal resource that strengthens the relationship between emotional contextual needs (e.g., a leader’s concern) and personal motives (i.e., learning) (e.g., Bai et al. 2016; Hameed et al. 2017). Specifically, individuals with strong emotional regulation that may interact with value-based leadership by easily understanding a leader’s styles better and obtaining stronger social exchange with and support from the leader (e.g., Tsai et al. 2016), eventually increasing dispositional and interpersonal motives to a large extent. Despite its important role in teams, emotional regulation has been rarely explored to verify its interaction with value-based leadership in team dynamics, which this study therefore examines.

This study differs from prior works in three crucial ways. First, it links value-based leadership to learning goal orientation and interactional justice so as to explain team performance in a single model setting, which has not yet been evaluated. Without a simultaneous evaluation of learning goal orientation and interactional justice as dual mediators, our understanding about the influence of value-based leadership on team performance will be highly limited, and managerial initiatives directed at developing effective leadership and improving team performance will turn out to be unjustifiable based on blind faith. For example, leaders wanting to improve team performance may mistakenly leverage learning goal orientation alone without reflecting upon whether they treat team members with respect or manage procedures fairly in their decision-making (i.e., interactional justice). Second, this work complements the literature by explicating the potential interaction of emotional regulation within team dynamics, which has been relatively understudied. While previous research has widely discussed the direct effect of emotional regulation on job satisfaction (Kohantorabi and Abolmaali 2014), learning behavior (Tsai and Chang 2014), or group performance (e.g., Günsel and Açikgöz 2013), how emotional regulation interacts with leadership styles to ultimately influence the development of team performance has been relatively understudied and is thus examined herein. Understanding the moderating role of emotional intelligence helps leaders prioritize different leadership foci to effectively guide different team workers. Third, this study contributes to the theoretical knowledge of value-based leadership by developing an inclusive model with the addition of learning goal orientation and interactional justice based on the social exchange theory. This study deepens our understanding of value-based leadership combined with the social exchange theory and provides a better explanation of team performance.

2 Research model and hypotheses

A modern approach that is able to instill a certain team culture and climate to motivate workers is termed value-based leadership. Value‐based leadership is an action‐based leadership style that can guide dispositional and interpersonal motives (Busch and Wennes 2012). The literature has suggested that leadership based on benevolent value (Karakas and Sarigollu 2013) and moral value (Hendrikz and Engelbrecht 2019) defines important value-based styles that dominate how a team performs. While benevolent leadership denotes a leader’s value that focuses on having individualized and holistic concern to team workers’ personal and familial well-being (Tan et al. 2016), moral leadership represents leaders’ styles driven by, and consistent with, an underlying set of moral value (Gu et al. 2020).

Drawing upon the value-based leadership theory, this study proposes a model (see Fig. 1) to elaborate how these two leadership styles influence team performance. In the model, team performance is indirectly influenced by benevolent leadership and moral leadership via two mediators that include interactional justice and learning goal orientation. At the same time, emotional regulation is hypothesized to moderate the respective effects of benevolent leadership and moral leadership on interpersonal justice and learning goal orientation.

Fig. 1
figure 1

Research model

Full size image

The goal orientation literature argues that a dispositional learning goal orientation strongly relates to team workers’ adaptability in order to effectively obtain the performance goal of complex tasks (Unger-Aviram and Erez 2016), because learning goals promote the growth of team workers’ professional knowledge and skills (Kozlowski et al. 2001) that help improve team performance. Hence, team workers who possess higher learning goal orientation are likely to achieve higher performance of a complex task (Seijts et al. 2004). At the same time, the organizational justice literature suggests that interpersonal treatment relates to performance outcomes (Fatima et al. 2015). According to the social exchange theory (Blau 1964), team workers who are fairly treated by their leader (i.e., interactional justice) are likely to reciprocate with positive behavior to effectively achieve team performance. For instance, empirical findings by Cohen-Charash and Spector (2001) have shown the positive relationship between interactional justice and in-role performance. Collectively, learning goal orientation and interactional justice both positively influence team performance.

The literature has indicated the positive effect of benevolent leadership on workers’ innovative learning behavior (Chan 2017; Gumusluoglu et al. 2016). Since a benevolent leader supports team workers with a quality social exchange relationship, the support given by a benevolent leader can substantially encourage team workers to exhibit outstanding learning tasks (Chan 2017). For that reason, team workers under benevolent leadership are likely to gain an enhanced sense of power to learn new things (Dedahanov et al. 2019), thus strongly focusing on learning goal orientation in order to perform better in the team. As a result, benevolent leadership can positively facilitate team performance via enhanced learning goal orientation, which is hypothesized below.

H1

Learning goal orientation mediates the positive relationship between benevolent leadership and team performance.

A moral leader shows respect to team workers and provides them autonomy for learning tasks and new skills that they carry out (Li et al. 2012), which in turn encourages their focus on developing their own competencies or skills (i.e., learning goal orientation) (Fasching et al. 2010). Previous research has argued that workers supervised by a moral leader are encouraged to perform learning tasks required for successful teamwork accomplishment (Avolio 1999; Masrukhin 2014). In other words, team workers under moral leadership are more willing to learn new knowledge and seek out opportunities that facilitate new skills based on their leader’s constructive feedback (Dedahanov et al. 2019), consequently improving their own creativity and performance (e.g., Dedahanov et al. 2016). Accordingly, the second hypothesis is now derived.

H2

Learning goal orientation mediates the positive relationship between moral leadership and team performance.

According to the value-based leadership theory, the values of kindness, benignancy, and altruism demonstrated by a team leader are likely perceived by team members as forming interactional justice (Della Corte et al. 2017; Wu et al. 2012). A benevolent leader expresses interest in team workers’ welfare, rewards them for desirable behavior, and helps them when they encounter difficulties (Niu et al. 2009). By showing goodwill and nurturance to team workers, a benevolent leader creates a supportive work environment whereby team workers experience respect, honesty, and genuine care from the leader, all of which are primary components of interactional justice (Gumusluoglu et al. 2020). In other words, a leader’s benevolent style represents his/her sensitivity to team workers’ needs, which drives interactional justice and eventually improves performance outcomes (Wu et al. 2012). The relationship between benevolent leadership and interactional justice can be explained through the rationality principle of the social exchange theory (He et al. 2017), in which fair treatment by a team leader in everyday encounters is rationalized, thus strengthening workers’ perceived interactional justice over time. To sum up, the hypothesis regarding the relationship between benevolent leadership and interactional justice is derived below.

H3

Interactional justice mediates the positive relationship between benevolent leadership and team performance.

According to the value-based leadership theory, moral leadership mainly utilizes values as the way of influencing team members (Bao and Li 2019). If team leaders behave with words and deeds that reflect the values they hold, then team members are likely affected to the extent that these values are assimilated (Bao and Li 2019). Moral leadership has been found to be useful in eliciting team effectiveness (Chen et al. 2015) due to two major reasons. First, a moral leader values workers’ inputs and encourages them to voice opinions (Brown et al. 2005), thus improving interactional justice. Such interactional justice in turns increases solidarity in teams and thus collective performance (e.g., Luo 2007). Second, a moral leader who treats team workers in an ethical manner is likely to foster a fair and just work atmosphere (Bao 2019), which motivates their proactivity to achieve team performance. In other words, a moral leader with the value of moral consideration often leads team members by the examples of upholding integrity, righteousness, and sensitivity towards team members, making the members perceive a strong feeling of justice toward leader-member interactions within the team (Wu et al. 2012). In summary, a leader who emphasizes moral standards by demonstrating ethical virtues and self-discipline is likely to treat workers fairly and sincerely (i.e., increased interactional justice) (Wu et al. 2012), consequently inducing their reciprocity to execute teamwork conscientiously. The hypothesis regarding moral leadership and interactional justice is thus derived as below.

H4

Interactional justice mediates the positive relationship between moral leadership and team performance.

2.1 Moderation of emotional regulation

Since team workers with emotional regulation abilities have a desire to develop high quality leader-member relationships (Adigüzel and Kuloğlu 2019) and are capable of appreciating positive leadership styles in ways that lead to positive interpretations of the workplace environment (George 2000), team workers with high emotional regulation tend to easily understand and appreciate benevolent leadership that offers sincere care to workers. When a benevolent leader demonstrates what he/she always does regarding individualized care to team workers (Lin et al. 2018a, b), team workers with high emotional regulation often become highly involved in teaming experiences and take proactive measures to adjust themselves and strengthen their dispositional and interpersonal motives (Carmeli 2003). In other words, team members’ emotional regulation facilitates the cognitive processes involved in learning and understanding their leader’s style (Cruz et al. 2020) that shows interpersonal concern and the willingness to do good to them beyond a selfish and egocentric motive (Nguyen 2010). As a result, team members with higher emotional regulation are more likely to practice learning goal orientation and perceive interactional justice. Collectively, benevolent leadership enhances learning goal orientation and interactional justice more strongly among teams with higher emotional regulation than those with lower emotional regulation. Two hypotheses are thus derived as below.

H5

Emotional regulation moderates the positive relationship between benevolent leadership and learning goal orientation such that the relationship is stronger when emotional regulation is higher.

H6

Emotional regulation moderates the positive relationship between benevolent leadership and interactional justice such that the relationship is stronger when emotional regulation is higher.

Previous research has indicated that workers’ emotion influences how they frame their interpretation of ambiguous circumstances (Schachter and Singer 1962). If team workers can regulate their negative emotions in teaming contexts, then they are more likely to shape the collective optimism (Quoidbach and Hansenne 2009) that can avert negative perceptions (or side effects) about moral leadership (e.g., excessive preaching or the uttering of pieties). On the contrary, when team workers are unable to self-manage their emotions, they are more likely to change their ideas and see things from a negative perspective. As a result, the good intention of moral leadership is likely disparaged by team workers with poor emotional regulation who tend to see things with a negative charge (Zhang et al. 2012). Therefore, the positive effects of moral leadership on learning goal orientation and interactional justice are likely weakened among teams with poor emotional regulation, leading to the next hypotheses.

H7

Emotional regulation moderates the positive relationship between moral leadership and learning goal orientation such that the relationship is stronger when emotional regulation is higher.

H8

Emotional regulation moderates the positive relationship between moral leadership and interactional justice such that the relationship is stronger when emotional regulation is higher.

3 Methods

3.1 Subjects and procedures

This study conducted empirical examinations using data from work teams across large insurance companies in Taiwan. Such work teams are considered to be appropriate sample for this study, because leadership and team performance are highly valued in the insurance companies. Researchers initially approached their EMBA alumni who worked as managers in the banking and insurance industry to assist in collecting survey data. In total, nine large nationwide insurance companies were investigated, and an anonymous survey was implemented. Research participants who voluntarily participated were assured that their responses would be only aggregated for statistical analyses, and thus no personal data would be individually disclosed.

This study applied a split-team sampling approach (Jiang et al. 2016) by collecting data from two different sources. A team leader and four members in each team were surveyed to measure different research variables. Team members measured five variables: team performance, learning goal orientation, interactional justice, benevolent leadership, and moral leadership. At the same time, team leaders measured team performance and emotional regulation. The literature has suggested the necessity of having team members and their leader jointly gauge team performance, because such performance represents a synthetic outcome that should be evaluated by members and their leader together to obtain an all-round objective assessment (Chiu et al. 2018). The literature has strongly recommended that research data be collected from different sources (e.g., members vs. leaders), because measuring different variables by different research subjects provides an advantage of alleviating common method biases (Lin et al. 2020).

Of the 500 questionnaires dispatched to 100 teams (i.e., 10–12 teams from each of nine insurance companies), 367 valid returned questionnaires from 79 teams in total were collected for a team-level response rate of 79%. The total of 79 team leaders comprised 49 male leaders (62.03%), 70 leaders who where 36 years or older (88.61%), and 56 leaders with work experience of 11 years or more (70.89%). In total, the 288 team members included 116 male workers (40.28%), 116 workers who were 36 years or older (40.28%), and 108 workers with work experience of 11 years or more (37.5%). Since it is a common practice that insurance organizations in Taiwan widely use small teams across hierarchical business tasks, 90% of the sample teams investigated by this study were small with the number of team members ranging from 3 to 5 (without a leader included). Therefore, with one team leader in each team, the total number of people in each team ranges from 4 to 6. In terms of the response rate from each team, this study obtained data from 11 teams with the rate between 50 and 79%, 11 teams with the rate between 80 and 99%, and 57 teams with the rate of 100%. Any teams with initially less than a 50% response rate were repeatedly contacted by researchers to eventually obtain sufficient data for analysis.

3.2 Measures

The variables in this work were measured with five-point Likert-type items modified from the literature (see Appendix). Before its actual survey, this study formed a focus group that compiled a research instrument and then conducted two pilot surveys to verify the quality of the instrument. The data from the second pilot survey were analyzed with exploratory factor analysis to verify the validity of the survey instrument. The research subjects in the two pilot surveys were excluded from the actual survey.

This study follows precautionary measures in the literature to minimize the potential threat of common method variance (CMV) (Liu et al. 2019). First, this study conducted an anonymous investigation to lessen subjects’ hesitation to fill out the questionnaires. Second, the measurement items in this study were repeatedly refined and improved by a focus group in terms of their readability and understandability. Third, different variables were measured by different subjects (i.e., from different sources).

This study calculated intraclass correlation to confirm the appropriateness of averaging individuals’ responses into team-level data. In Table 1, all ICC1 values were larger than the recommended level of 0.12 (James 1982). Second, two ICC2 values were larger than the recommended level of 0.60, but three values were not (Baruch and Lin 2012). Third, all rwg values were larger than the recommended level of 0.70 (James et al. 1984). Although some ICC2 values were smaller than 0.60, the literature has suggested that rwg values greater than 0.70 and ICC1 values exceeding 0.12 are considered quite sufficient and more important to warrant data aggregation (Rodríguez-Sánchez et al. 2017), because ICC2 is relatively more sensitive to a small number of team workers.

Table 1 Inter-rater reliability of the data from members
Full size table

It is important to note that the moderator measured by team leaders is considered team-level data, because they are the key people who can easily observe and evaluate the overall emotional regulation from a collective viewpoint due to their daily contacts with each member individually in the same team. The literature has provided a similar example like CEO pay that may seem like an individual-level variable, but in fact, it should be a firm-level variable, because every firm has only one CEO pay that represents its organizational strategy about incentives and compensation (Bhattacherjee 2012). From a theoretical perspective, team-level emotional regulation refers to the efforts of teams (observed by team leaders) to manage and express the emotions of team members in teamwork processes (Günsel and Açikgöz 2013; Reus and Liu 2004). In summary, this study focuses on team-level research based on all team-level data without cross-level issues.

Using two datasets of team members and leaders, CFA was performed separately. The test results in Tables 2 and 3 indicate that the figures of NNFI, CFI, and Bollen Non-normed Index Delta2 are larger than 0.9. The figures of RMR are smaller than 0.05, whereas the figures of RMSEA are slightly higher than 0.08. As most of the indices meet the recommended thresholds, we consider that the specified model matches the empirical data.

Table 2 Team-level standardized loadings of the data from 288 members across 79 teams
Full size table
Table 3 Team-level standardized loadings of the data from 79 team leaders
Full size table

Convergent validity was obtained based on three recommended criteria (see Tables 2 and 3): (1) factor loadings are significant at p < 0.001, (2) the figures of AVE (average variance extracted) are larger than 0.50, and (3) the figures of Cronbach’s alpha are larger than 0.70. At the same time, discriminant validity is obtained by chi-square difference tests. Since our chi-square difference statistics for all pairs of variables in Tables 4 and 5 meet the overall significance level to 0.01 or lower, discriminant validity is obtained.

Table 4 Chi-square difference tests of the team-level data from members
Full size table
Table 5 Chi-square difference tests of the team-level data from leaders
Full size table

3.3 Testing of hypotheses

This study conducted the analysis of hierarchical moderated regression to test its hypotheses. To reduce the possibility of unexpected biases, this study included relevant team-level control variables such as team leaders’ age, tenure, and so forth. Bernerth et al. (2018) suggested that control variables be included based on a reasonable explanation or evidence that indicates their relevance to focal variables in research. Therefore, this study controls variables relevant to team dynamics and performance, including team leader’s age (e.g., Kearney 2008), tenure (e.g., Fritz and Ibrahim 2010), familiarity (e.g., Lin and Chen 2000), education (e.g., Weisberg 1996), seniority (e.g., Tortorella et al. 2019), and social desirability (e.g., Densten and Sarros 2012). Table 6 shows the statistical results.

Table 6 Team-level hierarchical regression analysis
Full size table

In Model 1 the test result showed the significant effects of benevolent leadership and moral leadership on learning goal orientation with the coefficients of 0.51 (p < 0.01) and 0.39 (p < 0.01), respectively. In Model 2 this study revealed the significant effects of benevolent leadership and moral leadership on interactional justice with the coefficients of 0.56 (p < 0.01) and 0.32 (p < 0.01), respectively. In Model 3 this study found the significant effects of interactional justice and learning goal orientation on team performance with the coefficients of 0.53 (p < 0.01) and 0.30 (p < 0.01) respectively. In Model 4, two antecedents and two mediators were all included to explain team performance. The results revealed that the learning goal orientation and interactional justice in Model 3 remained significant in Model 4, whereas the effects of benevolent leadership and moral leadership on team performance were not significant. This phenomenon suggests no direct relationships between benevolent leadership and team performance and between moral leadership and team performance. To summarize the above results, H1–H4 are all supported.

To further test hypothesizes the moderating effects, this study introduced emotional regulation and its interaction terms with benevolent leadership and moral leadership in Models 5 and 6. First, emotional regulation shows a negative moderation on the relationship between benevolent leadership and learning goal orientation in Model 5 (thus, H5 is not supported). Second, emotional regulation does not moderate the relationship between benevolent leadership and interactional justice in Model 6 (thus, H6 is not supported). Third, emotional regulation positively moderates the relationship between moral leadership and learning goal orientation with the regression coefficient of 0.64 (p < 0.01) in Model 5 (thus, H7 is supported). Fourth, emotional regulation positively moderates the relationship between moral leadership and interactional justice in Model 6 with the regression coefficient of 0.63 (p < 0.01) (thus, H8 is supported). The analytical results of H1-H8 are all summarized together in Table 7.

Table 7 Empirical results of hypotheses
Full size table

To support the hypothesized mediation of learning goal orientation and interactional justice, this study performed bootstrapping analyses with 10,000 subsamples to double validate the mediation (see Table 8). The results confirm the mediating roles of learning goal orientation and interactional justice, because 95% confidence intervals do not cover zero.

Table 8 The results of the mediation using bootstrapping
Full size table

To verify the potential threat of CMV and CMB (common method bias), this study performed the methodology of ULMC (Williams et al. 1989) by comparing three models: (1) the trait model (i.e., the base CFA model in the preceding analysis), (2) the method model (with the unmeasured latent method construct), and (3) the restricted CFA model of ULMC (in which trait correlations and trait factor loadings are fixed with the values obtained from the base model). The results of ULMC are listed in Tables 9 and 10. The significant differences of the Chi-square fit statistics between Model 1 and Model 2 (see Tables 9 and 10) reveal the existence of common method variances. However, the insignificant differences of the Chi-square fit statistics between Model 2 and Model 3 (with Δχ2 (37) = 3.44 and Δχ2 (10) = 1.84 respectively in Tables 9 and 10) indicate that common method variances do not cause significant estimation biases (i.e., CMB) (Richardson et al. 2009; Williams et al. 1996). Accordingly, our empirical results are not substantially influenced by CMB.

Table 9 ULMC analysis for the team-level data from members
Full size table
Table 10 ULMC analysis for the team-level data from leaders
Full size table

4 Discussion

This study presents the moderating mechanism of emotional regulation and the mediating mechanism of learning goal orientation and interactional justice from the perspective of leadership styles. This study reconciles the arguments regarding what value-based leadership styles are critical and whether dispositional and interpersonal motives mediate the development of team performance. On the basis of its findings, this study contributes to the literature by providing the following theoretical and managerial implications.

4.1 Theoretical implications

This study offers three theoretical implications in particular. First, it justifies the positive effect of benevolent leadership on interactional justice, which is analogous (but not theoretically identical) with the social influence theory that considers leaders as the authority of social principles of fairness in shaping the justice climate in business firms (Gumusluoglu et al. 2020). More specifically, although the literature applies such climate perception as social influence that facilitates interactional justice (Gumusluoglu et al. 2020), this study thus offers a social exchange perspective that justifies the formation of interactional justice. This study also finds that benevolent leadership enhances team performance through increased justice perception, which addresses the recent calls in the human resources discipline for a better understanding of mediating mechanisms by which benevolent leadership increases or reduces team performance (e.g., Li et al. 2018).

Second, the positive effect of moral leadership on learning goal orientation in this study is consistent with (but theoretically distinct from) the previous argument based on the social learning theory (Moss et al. 2019). While previous literature uses the social learning theory to theorize that ethical leadership is both normative and worth at emulating subordinates to eventually intensify their felt obligation (e.g., learning goals) (Moss et al. 2019), this study offers a reciprocal aspect that triggers the development of learning goal orientation. It integrates the traditional discussion of moral leadership into new theoretical territory—the collective moral duty of self-improvement based on learning goal orientation—and thus complements the literature that often focuses on typical outcomes of moral leadership, such as moral behavior, organizational citizenship behavior, and deviant behavior.

Third, the positive moderation of emotional regulation in this study provides another explanatory route to complement the conservation of the resource theory arguing that workers strive to preserve, retain, protect, and develop valued resources by simultaneously mitigating any threats of resource loss (Hobfoll 1989). In this study, team workers with stronger emotional regulation may have sufficient resources to precisely follow moral leadership and to enhance their valued motives such as interactional justice and learning goal orientation, consequently achieving great team performance. This study recognizes that value-based leadership may be highly limited without emotional regulation taken into account. This study contributes a novel model that enhances our understanding of the simultaneous importance of emotional regulation and value-based leadership in teamwork processes. Complementing what has been learned about value-based leadership in the literature, this study illustrates that when levels of emotional regulation are low, leaders should provide supportive resources (e.g., mindfulness training, emotional therapy) as the first priority so that their value-based leadership can effectively work.

4.2 Managerial implications

This study offers managerial implications for teamwork practices. To begin with, the positive effect of benevolent leadership on learning goal orientation and interactional justice suggests that team leaders can learn to display holistic concern for individuals’ workplace incidents and personal life events in order to increase learning goal orientation and interactional justice. Such concerns should be accompanied by a sincere affective attitude to avoid suspicion or misunderstanding. At the same time, team leaders should develop strong observational skills so as to sense individuals’ needs and then provide necessary support in a timely manner.

The positive effect of moral leadership on learning goal orientation and interactional justice suggests that team leaders show great perseverance in terms of self-discipline, virtues, moral principles, and unselfishness in order to encourage learning goal orientation and promote interactional justice effectively. Such perseverance may also minimize the risk of unethical behavior performed by workers, because the excessive pursuit of goal orientation may lead to increased unethical behavior (Welsh et al. 2019).

The positive moderating effect of emotional regulation reveals that moral leadership is more important to team workers when emotional regulation is high. In other words, team leaders should understand that their moral leadership can become less influential if the team’s emotional regulation is poorly fostered. To effectively leverage the influence of their moral leadership, leaders should encourage team workers to observe vocal emotional cues within teams and reflect upon what they can do creatively in order to avoid any emotional overreaction, consequently maximizing the positive effect of moral leadership.

In conclusion, it is not easy to achieve high team performance with merely a subjective degree of management. An effective strategy to improve team performance is to provide team leaders with appropriate leadership training and education regarding when and how to implement benevolent and moral leadership by simultaneously detecting the team-level emotional regulation. By making good use of the mediating and moderating mechanisms explored in this study, team leaders are likely to tailor sound teamwork tactics and actions to obtain satisfactory team performance.

4.3 Limitations and future research

There are two major limitations in this study that imply future research directions. The first limitation is its generalizability of research findings in teaming contexts due to the sample teams from insurance companies only. Therefore, the empirical findings might not be highly generalizable to workgroups across different industries. Second, due to its theoretical foundation based on the social exchange theory and value-based leadership theory, this study does not address cultural or political variables (e.g., uncertainty avoidance, masculinity, politics, opportunism) to clarify team performance. For example, social theories such as the situated cognition theory, rational action theory, power-political theory, or socio-cultural theory may be integrated with the value-based leadership theory to explain team performance from a more comprehensive standpoint. Future investigation on a wide variety of industries across different countries can increase both research generalizability and impact in terms of managerial science. Researchers can thus investigate more diverse work teams with longitudinal observations, integrate different theories to justify team dynamics, and explore a wide variety of moderators so that useful strategies of managing work teams can be exhibited.