Page 199 - 33-3

NTU Management Review Vol. 33 No. 3 Dec. 2023

suggesting that our findings are robust.

4. Research Limitations

There are several limitations needed to be noted. First, the mean and standard
deviation of abusive supervision are quite low, which may cause the problem of range
restriction. Second, we did not include contextual variables in our model (e.g., political

climate), which might also influence subordinates’ evaluations of leaders’ abusive
behaviors. Finally, we collect the ratings of abusive supervision and leader effectiveness
only from subordinates, which increases to the possibility of Common Method Variance
(CMV). Nonetheless, we did collect abusive supervision and leader effectiveness
at different time points and the correlation between abusive supervision and leader

effectiveness is relatively low. Thus, the CMV problem should not influence our findings
adversely.

5. Originality/Contribution

The present study attempts to make the following contributions to the abusive
supervision literature. Theoretically, we propose an integrative model to understand the
boundary conditions of the abusive supervision-leader effectiveness relationship, and find

that abusive leaders with high physical attractiveness can increase subordinates’ perceived
leader effectiveness, which in turn increases subsequent supervisor-directed helping.
However, practically we are not suggesting that leaders should engage in abusive behaviors
to shape subordinates’ evaluations. Rather, the results of our study clarify “what kind of

skills or characteristics” can influence subordinates’ reaction to abusive supervision. Thus,
organizations should pay more attention to the abusive behaviors displayed by leaders with
low political skills and high physical attractiveness. Empirically, we control for negative
emotions and LMX in our analyses to rule out alternative mechanisms, suggesting that our

proposed political activity perspective can meaningfully explain the abusive supervision-
leader effectiveness above and beyond the affective and relational mechanisms.

191