In the recent AMJ (48-3, pp. 420-432) Kenny Law and his colleagues shows that
LMX leadership completely mediates Transformational leader behavior. When
the contribution of LMX leadership is partialled out, Transformational behavior
is not related to either job performance or organizational citizenship. But,
LMX leadership is related to both of these variables when transformational
behavior is partialled out.
The model these results support is that the primary predictor of both
performance and citizenship is LMX leadership and the secondary predictor of both is
transformational behavior. With a weak LMX bond, transformational behaviors
make no real contribution, but with a strong LMX bond, leaders style makes a
contribution. Thus, the primary and secondary goals of leadership are clear:
first, make the LMX bond authentic and second adopt a supportive style.
Clearly, the first is much more difficult and is frequently denied. How do we get
leadership scholars to accept that leadership demands a strong LMX interpersonal
bond?