EV-0045
What gets you the job is not what makes you good at it
Grijalva, E., Harms, P. D., Newman, D. A., Gaddis, B. H., & Fraley, R. C. (2015). Personnel Psychology, 68(1), 1-47 · Source
What this entitles us to say
A meta-analysis finding that narcissism has a positive relationship with leadership emergence - who gets chosen to lead - but no linear relationship with leadership effectiveness, and that the apparent null conceals a curvilinear relationship in which moderate levels do best.
What it found
Meta-analytic integration of the narcissism-leadership literature, testing linear and nonlinear relationships:
- Leadership emergence - who is selected, who rises: positive relationship with narcissism. - Leadership effectiveness - how well they lead once there: no linear relationship. - The null on effectiveness masks a curvilinear trend - moderate levels relate best, with too little and too much both worse.
Why it matters here
If The Case for Character needs one finding, this is it.
Organizations select for confidence, presence and self-promotion. Those characteristics load onto something that reliably produces emergence and does not produce effectiveness. Which means the selection system is not merely imperfect - it is systematically rewarding a thing that does not predict the outcome it is meant to predict. That is not a moral complaint about executives; it is a measurement failure with a meta-analysis behind it.
It also sets up the program's own position precisely. The book's argument is that character is a choice available to everyone while skill is not, which is why you interview for skill and set a standard for character. This finding adds the missing piece: the qualities that win the interview are the least diagnostic ones in the room.
Two guardrails for using it. First, never point it at a person - the fourth principle bars clinical vocabulary in participant-facing material, and "narcissist" is the most abused word in that category. Second, the curve matters: this is not "confidence is bad." It is "the amount that gets you promoted is past the amount that helps."
The limits
Narcissism is a personality construct measured by questionnaire, not a synonym for arrogance or bad character, and this program must never use it as a label for a person - clinical terminology in participant-facing material is forbidden by the fourth principle for exactly this reason. Leadership effectiveness here is rated by others, so this compares perceptions. Correlational. The curvilinear finding is a modeled pattern rather than a threshold anyone should apply to an individual. Verified to abstract level.
What we do not say
"Narcissists are bad leaders," or anything that applies this to a named person. It is a population-level relationship using a questionnaire construct, effectiveness is rated by others, and the program does not diagnose people.