53 / 274
臺大管理論叢
第
27
卷第
3
期
53
future research might adopt an experimental or longitudinal design with multiple sources to
diminish the CMV effect. Moreover, with expection to the supervisor, colleague, and self-
report; the OCB examination also needs a third party to provide evaluations (Bolino and
Turnley, 2005).
Second, Chan (2013) empirically found out that benevolent leadership behavior did not
lead to employees’ voice behavior; however, our results did not appear to back him up. Voice
behavior is regarded as one of OCB (e.g., Podsakoff et al., 2011; Van Dyne and LePine,
1998); the result of Chan (2013) and our studies demonstrated the inconsistent outcome for
the relationship between benevolent leadership and OCB. Therefore, future research might
examine the relationships between benevolent leadership and various dimensions of OCB.
Third, although the results revealed that trust in organizations and in leaders mediated the
positive relationship between benevolent leadership and OCB, we did not separate OCB to
OCBO and OCBI (Lee and Allen, 2002; Williams and Anderson, 1991) as criterion variables
to respectively correspond to those two forms of trust. Future research should do so.
Fourth, the multilevel method has become an issue in leadership research; future
research might adopt hierarchical linear modeling to examine the influence of benevolent
leadership in a multilevel context (Lam, Huang, and Lau, 2012).
Finally, scholars are increasingly concerned about the self-interest and negative effects
of OCB (Bolino et al., 2013; Bolino and Turnley, 2005; Bolino et al., 2004). For instance, too
many behaviors outside one’s role in becoming in-role obligations might result in overloaded
and lower performance. Thus, the effect of various antecedents and moderators on positive
and negative OCBs should be discussed in future research.
5. Originality/Contribution
Our results, similar to previous reports (e.g., Chan et al., 2013; Chan and Mak, 2012),
show the positive effect of benevolent leadership on OCB. This positive relationship,
however, could be reversed if subordinates perceive high manipulative intention from a
leader. That is, benevolent leadership does not always lead to positive consequences.
Moreover, the interaction of benevolent leadership and perceived manipulative intention
influences OCB and this effect is mediated by trust in supervisor and organization.
This study responds to previous researches (e.g., Pellegrini and Scandura, 2008;
Pellegrini, Scandura, and Jayaraman, 2010) that pay more attention to empirical
examinations on Chinese leadership, which includes the benevolent leadership. It also