قدرت دادن به رهبری تیم و عملکرد ایمنی در نیروگاه های هسته ای : یک رویکرد چند سطحی

Despite the large body of work on team leadership, hardly any literature has dealt with team leadership in safety performance settings. The goal of the present study is to analyze how team leader behaviors influence team members’ safety performance in nuclear power plants. For this purpose, an empowering leadership approach was assessed. We consider a multilevel model in which safety performance is divided into three types of behaviors. The sample was composed of 479 workers in 54 groups from two Spanish nuclear power plants. The results suggested that leaders’ empowering behaviors generated higher safety compliance behaviors and higher safety participation behaviors by team members, whereas risky behaviors were reduced. Empirical support was found for hierarchical linear modeling linking leadership and safety performance behaviors. Practical implications, study limitations and directions for future research are discussed.

The role of leadership is of prime importance for the safety of nuclear power plants (NPPs). In general, leadership is viewed as a shift lever for safety culture, and as an important antecedent of achieving high levels of safety (e.g., Fahlbruch, 2005, International Nuclear Safety Advi, 2002, International Atomic Energy Agency, 2007 and International Atomic Energy Agency, 2008). Although safety leadership is promoted through seminars, simulator trainings or safety culture reviews in the nuclear industry, only a few studies have empirically analyzed and specified the impact of leadership on safety performance (e.g., Martínez-Córcoles et al., 2011 and Yule et al., 2007). The question of which leadership style might best fit within the highly regulated work context of nuclear power plants still remains unanswered. Whereas leadership theories are primarily concerned with enhancing the effectiveness and efficiency of employees’ work performance, it is questionable whether these promoted leadership styles will obtain the same positive results within a work context where the trade-off between efficient and safe performance is sometimes crucial. The present study aims to examine the differential impact of safety leadership on safety performance in that context. Using a multi-level approach, we tested the effects of an empowering team leadership style (ELQ, Arnold et al., 2000) on safety performance. As a result, we introduce three dimensions for the safety performance construct, namely, safety compliance, safety participation and risky behavior. In the next two sections, we introduce research on safety leadership and the leadership model utilized in our study.

The present study was designed to test the effect of team leadership on safety performance. Our goal was to test a model to explain the differential impact of team empowering leadership on three facets of safety performance (safety compliance, safety participation and risky behaviors) in nuclear power plants. Therefore, we expand the findings of a recent study by Martínez-Córcoles et al. (2011), in which it was shown that EL positively influenced safety behaviors by means of safety climate. Apart from this study, no other study has investigated the influence of empowering leadership on safety performance within the nuclear field. Two directions have been adopted in the present study. The first was to more fully explore the direct relationships between EL and safety performance from a multilevel approach (team EL). The second was to conceptualize safety performance as three different types of behaviors, with the novelty of adding risk-taking behavior to the predominant duality of safety performance (safety compliance-safety participation) in safety research. The results of HLM show clearly that leadership significantly predicts the three safety performance behaviors. In light of the findings from our study, we can conclude that team EL is a potential antecedent of safety performance. Specifically, when leaders behave like empowering leaders, they produce compliance with safety procedures and requirements (Hypothesis 1), enhance safety participation behaviors (Hypothesis 2), and they are able to reduce team members’ risky behaviors (Hypothesis 3). Thus, our study reveals that EL directly influences subordinates’ safety performance behaviors. Our results are in line with Zohar’s (2003) conclusion that behavioral leadership can directly influence employees’ behaviors (without mediators) because of the effects of social learning processes (Bandura, 1977). Regarding social learning theory, team leaders’ behaviors (i.e. leading by example) are observed by team members, who imitate these behaviors because they are rewarded by team leader or team colleagues (i.e. due to recognition of these behaviors). EL is a purely behavioral leadership style, and in our study we recorded the team members’ responses regarding the direct interaction with their leader. Thus, our study is a pioneer in finding direct relationships between leadership and safety performance, shedding empirical light about a new way in which leaderships proxies to transformational or participative styles (e.g. empowering leadership) impact on safety performance, beyond the already known mediated effect of safety climate mentioned in previous empirical studies (e.g. Clarke and Ward, 2006, Kelloway et al., 2006, Martínez-Córcoles et al., 2011 and Zohar, 2002). We believe that the results of our research could have important implications for safety research in two ways. First of all, a multilevel approach was used, which allows us to detect the group effects and the influences (social interaction, team membership, etc.) to which individuals are subjected. Therefore, using hierarchical linear modeling, we estimated three cross-level relationships, assuming dependent observations due to the hierarchical data structure. That is, individual shared perceptions within units about their team leader were considered. Second, measuring safety performance in HROs implies not only focusing on safety compliance and safety participation (behaviors associated with positive safety performance), but also considering behaviors such as deviance from rules and procedures that involve risks and could produce fatal consequences. In that case, we found that empowering leaders’ behaviors were related to a decrease in employees‘ risky behaviors. The latter result needs some more clarification, especially, when it comes to the practical implications of our study. As stated above, we argue that highly standardized systems (in terms of procedures and safety rules) like nuclear power plants substitute leadership functions by aligning the goals of organizational members to safety and reliability. Promoting an empowering leadership style in nuclear power plants implies that team members have more autonomy in fulfilling their work tasks, which might foster their risk-taking behavior. However, we think that the alignment of safety goals due to procedural systems paired with an empowering leadership styles counteracts this argument. Empowering leaders motivate their subordinates to critically question rules and procedures,4 which will be a much more harder endeavor for (e.g.) authoritarian leaders who will mainly focus on rule compliance without subordinate‘s critical questioning of the match between the rule and the situation. Thus, by fostering safety knowledge exchange (i.e. more safety participation) empowering leaders support their subordinates to better evaluate rule compliance situations, to make safety-optimizing decisions and to cope with unknown (or not anticipated) situations, since they have internalized the goal of to be developed behavior (i.e. safety) and therefore have the knowledge to cope with these situations in a safety-directed manner. Empowering leadership provides meaning to workers’ rule compliance in responding to complex problems. When people perceive that complying with rules makes sense, they get a better understanding of potential hazards and are motivated to develop their adaptive capacity further. The results of our study clearly suggest that this might be the case (e.g. less risk-taking behavior, more safety participation and more safety compliance). Several limitations exist in the present study. First, we use self-reports to measure safety performance. This may lead to inflated results due to respondents’ tendencies to respond in a consistent and socially desirable manner. Thus, we measured perceived safety behaviors and, therefore, do not know the real occurrence and outcome of these behaviors. Second, the cross-sectional character of the study reduces the nature of the variables to a “snapshot”, instead of being dynamic processes over time. For instance, how leadership develops its impact on each type of safety performance over time is missing, and causal relations cannot be tested. Third, we have not taken mediator variables into account. Most previous empirical research (although not particularly in HROs) agrees that safety climate is the mediator variable through which leadership influences safety outcomes (e.g., Zohar, 2000). However, one reason why we explored the direct relationships between leadership and safety performance is because of the recent appearance of the EL model (to date there is no empirical multilevel research about EL for safety). Therefore, first testing the direct cross-level relationships between team EL and safety performance types has been our main goal in this study. Another one was to test if a behavioral leadership style could have direct impact on employees’ safety performance as previously has been suggested (e.g., Zohar, 2003). Finally, the reader should be careful regarding the generalizability of the results of the present study. According to a cross-cultural comparison of work in reactor control rooms, Rochlin and von Meier (1994) found several functional variations, which can be lastly attributed to the impact of national cultures. At least, four different cultural archetypes varying on the dimensions “hierarchical power distance” and “rule orientation” were identified (Latin, Germanic, Scandinavian-British, and Asia countries). Therefore, our results concerning the outcomes of empowering leadership should be seen in the light of this impact. They could be altered depending on the cultural milieu embedded within the national culture. Future studies, which generalize data susceptible to be influenced by culture, should take this limitation into consideration. The limitations detected also indicate possible directions for future research. Empirical studies should explore potential mediators and/or moderators (i.e. group-level safety climate) in the team leadership-safety performance relationship from a multilevel approach. Moreover, future research should consider a longitudinal design in order to capture the dynamic quality of the team leadership construct over time. At the same time, other methods should be used in addition to self-report instruments, such as observations or interviews, to control tendencies of consistent responses or social desirability. The authors hope that the present study sheds new light on leadership and safety performance in HROs, and provides new orientations for practitioners.