Steven G. Hertz and Tobias Krettenauer
Review of General Psychology, Vol 20(2), Jun 2016, 129-140.
http://dx.doi.org/10.1037/gpr0000062
Abstract
This meta-analysis examined the relationship between moral identity and moral behavior. It was based on 111 studies from a broad range of academic fields including business, developmental psychology and education, marketing, sociology, and sport sciences. Moral identity was found to be significantly associated with moral behavior (random effects model, r = .22, p < .01, 95% CI [.19, .25]). Effect sizes did not differ for behavioral outcomes (prosocial behavior, avoidance of antisocial behavior, ethical behavior). Studies that were entirely based on self-reports yielded larger effect sizes. In contrast, the smallest effect was found for studies that were based on implicit measures or used priming techniques to elicit moral identity. Moreover, a marginally significant effect of culture indicated that studies conducted in collectivistic cultures yielded lower effect sizes than studies from individualistic cultures. Overall, the meta-analysis provides support for the notion that moral identity strengthens individuals’ readiness to engage in prosocial and ethical behavior as well as to abstain from antisocial behavior. However, moral identity fares no better as a predictor of moral action than other psychological constructs.
And the conclusion...
Overall, three major conclusions can be drawn from this metaanalysis. First, considering all empirical evidence available it seems impossible to deny that moral identity positively predicts moral behavior in individuals from Western cultures. Although this finding does not refute research on moral hypocrisy, it put the claim that people want to appear moral, rather than be moral into perspective (Batson, 2011; Frimer et al., 2014). If this were always true, why would people who feel that morality matters to them engage more readily in moral action? Second, explicit self-report measures represent a valid and valuable approach to the moral identity construct. This is an important conclusion because many scholars feel that more effort should be invested into developing moral identity measures (e.g., Hardy & Carlo, 2011b; Jennings et al., 2015). Third, although moral identity positively predicts moral behavior the effect is not much stronger than the effects of other constructs, notably moral judgment or moral emotions. Thus, there is no reason to prioritize the moral identity construct as a predictor of moral action at the expense of other factors. Instead, it seems more appropriate to consider moral identity in a broader conceptual framework where it interacts with other personological and situational factors to bring about moral action. This approach is well underway in studies that investigate the moderating and mediating role of moral identity as a predictor of moral action (e.g., Aquino et al., 2007; Hardy et al., 2015). As part of this endeavor, it might become necessary to give up an overly homogenous notion of the moral identity construct in order to acknowledge that moral identities may consist of different motivations and goal orientations. Recently, Krettenauer and Casey (2015) provided evidence for two different types of moral identities, one that is primarily concerned with demonstrating morality to others, and one that is more inwardly defined by being consistent with one's values and beliefs. This differentiation has important ramifications for moral emotions and moral action and helps to explain why moral identities sometimes strengthen individuals' motivation to act morally and sometimes undermine it.