Why leaders can be bad: Linking rigor with relevance using machine learning analysis to test the transgression credit theory of leadership

Abstract

Transgression credit is a form of deviance credit that occurs when people are more permissive towards transgressions by in-group leaders than by in-group nonleaders and out-group members and leaders. Despite rigorous experimental and simulation evidence for transgression credit, the ability to make such group processes research relevant to organizations and wider policy requires evidence with greater ecological validity. We examine transgression credit using spontaneously arising data from Twitter (now X) to test theoretically specified reactions to instances of transgressive leadership by the UK Prime Minister Boris Johnson. Studies 1a and 1b compared Conservative and Labour Members of Parliament’s (MPs’) tweets in response to Boris Johnson’s unlawful prorogation of Parliament (Study 1a) and his publication of an Internal Market Bill that would break international law (Study 1b) with tweets responding to a nonleader, Dominic Cummings, breaking coronavirus lockdown rules. Conservative, but not Labour, MPs were more permissive of Johnson’s, but not Cummings’, transgression. Study 2 examined the semantic themes occurring among supportive and unsupportive tweets posted by the UK general public in response to Boris Johnson’s unlawful prorogation of Parliament. Across studies, the evidence is consistent with propositions from deviance credit and social identity theories.