A Fatal Flaw: Why Measures of Authentic, Ethical, and Servant Leadership Spuriously Predict Objective Outcomes
Across three studies, we show that measures of three leadership styles—authentic, ethical, and servant leadership—predict objective outcomes, even if leadership is constant. That is, variation in such measures of leadership correlates with variation in the outcomes, although there is no variation in leader behaviors and characteristics. How are such consistent, yet seemingly nonsensical results possible? One answer might be: measurement error. And indeed, the leadership style measures are erroneous. However, such an error is anything but random and instead follows a systematic pattern. In our studies, we demonstrate that the allegedly behavioral measures of leadership are themselves non-behavioral outcomes. They represent general subjective evaluations of a leader, which are influenced at least by business success, moral achievements, and leader-follower value alignment. Therefore, predictions that leadership styles impact objective outcomes are spurious and simply mean that subjective outcomes—ratings of leadership styles—correlate with other objective outcomes. We further support this claim by developing and using a new measure: the so-called meaningless leadership questionnaire. This questionnaire consists of such non-behavioral and vaguely positive items, like “Is an interesting human being” or “Has a unique character”. In Studies 2 and 3, we show that this questionnaire has not only fine psychometric properties, but basically the same predictive properties as authentic, ethical, and servant leadership. Such a finding further suggests that measures of the three leadership styles are rather measures of vaguely positive rater evaluations than of concrete leader behaviors. Thus, we should doubt the validity of much existing research and instead develop new measures of leadership behaviors.