their subordinates. PL impacts follower outcomes (in-role performance,
OCB, and work attitudes) through followers’ cognitive-motivational
states (fear, gratitude, and identification). As we noted in the previous
section, this mediated relationship has been substantiated by prior
research. There is also considerable evidence that traditionality and
resource dependence among followers moderates the relationship
between PL and subordinates’ psychological responses or states.
The upper-level and lower-level leadership processes interact in
three ways in the multilevel model. First, there is a top-down influence
flowing from upper-level leaders to lower-level leaders. By virtue of
their high-status position in the hierarchy and their ability to control
resources, high-level managers are natural role models for low-level
managers (Antonakis and Atwater, 2002 ; Bandura, 1996). This role
modeling effect is likely to be especially strong when the leadership/
management style of upper-level managers is perceived as effective by
lower-level managers.
Second, the paternalistic management style of upper-level man-
agers and the organizational climate will impact the followers’
cognitive-motivational states directly beyond the influence of the
low-level managers’ PL. In a society with high power distance and
collectivism, the paternalistic management practices/climate will
not only shape subordinates’ cognitive-motivational states, but also
may serve as a situational enhancer for the lower-level PL effects.
As Schneider ( 2000 ) has pointed out, the organizational climate could
signal the management’s expectations to employees and help them
to determine which behavior is appropriate in a given work environ-
ment. Employees in a work unit where the paternalistic climate has
been established are more likely to be receptive to their direct super-
visor’s PL. Thus, we expect a synergy between low-level managers’ PL
and a paternalistic climate in the work unit. Besides the top-down
effects, there is a bottom-up effect in the model as well. Outcomes
at the individual level in aggregate will influence outcomes at the
unit level.
In summary, this multilevel approach to PL corroborates the
recommendation to examine the impact of leadership behaviors from
both the perspective of individual employees’ idiosyncratic experi-
ences, and the shared view of work-unit members’ overall experience
(e.g. Dansereau and Yammarino, 1998 ; Liao and Chuang, 2007 ).
We propose not only differential mechanisms for upper-level and
Paternalistic leadership in Chinese organizations 199