Secure and defensive types of high self-esteem: Discrepancies between explicit and implicit self-esteem predict defensiveness

Longstanding theories suggest that high self-esteem can assume qualitatively different forms that are related to defensiveness. These theories propose that high self-esteem accompanied by negative self-feelings that exist at less conscious levels may be particularly defensive. The present research empirically tests this possibility by using a new measure of implicit self-esteem, the Implicit Association Test (IAT), to operationalize less conscious self-feelings—implicit self-esteem represents highly efficient evaluations of the self that are believed to exist largely outside of awareness. Thus, we predicted that individuals who report a high level of (explicit) self-esteem, but evidence a low level of implicit self-esteem will behave more defensively than individuals who are high in both explicit and implicit self-esteem. In Study 1, we demonstrate that the correspondence between explicit and implicit self-esteem predicts levels of narcissism, an individual difference variable known to be closely related to defensiveness. Individuals with high explicit but low implicit self-esteem showed the highest levels of narcissism overall. Study 2 extends this finding by showing that participants with high explicit self-esteem but low implicit self-esteem display the most dissonance reduction in the classic free-choice dissonance paradigm—a defensive reaction. Next, Study 3 shows that among participants with high explicit self-esteem, there is a negative relation between implicit self-esteem and in-group bias in the minimal group paradigm. Participants with high explicit but low implicit self-esteem favoured fellow in-group members more than did participants with high explicit and implicit self-esteem. Studies 4a and 4b examine a similar process between existing (rather than minimal) groups; that is, racial discrimination. Participants had their self-views threatened with negative feedback, and then indicated the severity of punishment a student deserved for starting a fistfight. Among participants with high explicit self-esteem (i.e., in the top third of the distribution), in both Study 4a and 4b, those with low implicit self-esteem recommended a more severe punishment for a Native student offender than for a White offender. Thus, individuals with high explicit but low implicit self-esteem may be most prone to use racial discrimination as a means of protecting threatened self-views. Across these studies, we found convergent evidence that high explicit self-esteem can assume secure or defensive forms that are related to less conscious self-feelings. High explicit self-esteem accompanied by low implicit self-esteem may represent a form of defensive self-esteem, whereas high explicit self-esteem accompanied by high implicit self-esteem may represent a form of secure self-esteem.