(Farhet al., 2006 ). This is not surprising given the centrality of the
norm of reciprocity (bao) in Chinese societies (Yang, 1957 ). While
some traditional values (e.g. respect for authority, male domination)
have been eroded by societal modernization, the reciprocity value
endures (Cheng and Farh, 2001 ; Yang, 1998 ). As expected, benevolent
leadership also had positive effects on work attitudes and performance
including satisfaction with the leader, loyalty to the leader, commit-
ment to the organization, job performance, and OCB (Cheng, Huang,
and Chou, 2002 ; Cheng, Shieh, and Chou, 2002 ; Chenget al., 2003 ;
Chenget al., 2004 ;Farhet al., 2006 ).
Main effects of moral leadership
Consistent with the prediction of Farh and Cheng’s ( 2000 ) model,
moral leadership was strongly associated with the subordinate’s psycho-
logical response of identification with the leader (Chenget al., 2004 ;
Farhet al., 2006 ). When all three PL dimensions were entered into the
regression equation simultaneously to predict identification with
the leader, moral leadership had the largest beta weight (Farhet al.,
2006 ). Similar to benevolence, moral leadership was found to be
consistently positively correlated with subordinate outcomes such as
satisfaction with the leader, loyalty to the leader, commitment to the
organization/team, job performance, and OCB (Cheng, Huang, and
Chou, 2002 ; Cheng, Shieh, and Chou, 2002 ; Chenget al., 2003 ;
Chenget al., 2004 ;Farhet al., 2006 ). In the above analyses, moral
leadership was often found to have the largest effect on subordinate
outcomes (Chenget al., 2003 ; Farhet al. 2006 ), and, as pointed out
earlier, its positive effect on compliance was even stronger than the
effect of authoritarianism (Chenget al., 2004 ; Farhet al., 2006 ). These
findings point to the significance of the leader’s morality in leader–
subordinate relationships in contemporary Chinese organizations.
Interactive effects of the three dimensions
Since each of the three dimensions of PL captures a key aspect of
paternalistic leadership, a logical research question follows: do they
mutually reinforce each other in producing interactive effects that go
beyond their individual main effects? Although Farh and Cheng
( 2000 ) were mute on the three-way interactions, they did point out
that high authoritarianism in conjunction with high benevolence
represents an ideal type of leadership widely accepted in traditional
178 Jiing-lih Farhet al.