The vaccination rift effect provides evidence that source vaccination status determines the rejection of calls to get vaccinated.

COVID vaccination protects individuals and helps end the pandemic, but a sizable minority in Western countries rejects the vaccine. Vaccination status should serve as a group membership, critical communication between groups undermines trust, and we accordingly suggest that calls to get vaccinated by vaccinated sources lead to defensive rejection instead of desired behavior change. We term this the vaccination rift effect. A unique collaboration with national print, online and TV news media yielded a large (N = 1170), age-representative sample of Austrian citizens for our fully randomized experiment. Participants exhibited the vaccination rift: They ascribed less constructive motives, d = 0.28, 95% CI [0.17; 0.40], experienced more threat, d = - 0.30, 95% CI [- 0.42; - 0.19], and ascribed worse personality characteristics to vaccinated (vs. unvaccinated) commenters, d = 0.17, 95% CI [0.06; 0.29]. Constructiveness consistently predicted behavioral measures of counterarguing and vaccination planning (indirect effects B = 0.033, SE = 0.013 and B = - 0.056, SE = 0.014). The vaccination rift was substantially stronger among the critical group of unvaccinated participants, ds = |0.39-0.52|, than among those fully vaccinated, ds = |0.08-0.17|. We discuss how to apply these psychological mechanics of the vaccination rift to public campaigns.