Children grade teachers on consistency, not severity: 93% demanded equal punishment

When two story characters broke the same rule and received different consequences, 93% of the children asked said the teacher should change something; when both received the same consequence, whether punished or let off, 97% said nothing needed changing ^[5][6]. The study, published in the Journal of Experimental Child Psychology by Katarzyna Myslinska-Szarek of SWPS University and Felix Warneken of the University of Michigan, points at classroom discipline as a consistency problem rather than a deterrence problem ^[2][3].

The design was ordinary on purpose. A total of 122 children aged 6 to 9 in Poland and the United States were shown picture stories in which characters used smartphones at school or left a mess after lunch, then answered questions including whether the character should be punished and whether the punishment was fair ^[1].

The mechanism the team had already described in Cognition is a primacy effect: the adult's decision about the first child who breaks a rule sets the reference point for everything after it ^[4]. Punish the first offender and children expect the second to get the same; let the first one go and they expect the second to be let go too ^[4]. The researchers attribute this to meta-norms, rules about how rules must be applied - impartially, and to everyone ^[8]. That is not a young-child artefact of not yet understanding proportionality; even preschoolers expect antisocial behaviour to be punished, and older children already grasp that punishment should scale with the offence ^[12].

The two headline percentages are worth sitting with. Demand that the teacher change something rose from roughly 3% under symmetric treatment to 93% under asymmetric treatment, a swing of about 90 percentage points driven entirely by whether the two children were handled alike ^[14]. "Being treated differently by an adult provokes strong resistance and undermines their sense of justice," Myslinska-Szarek said of the results ^[7].

The emotional accounting runs the same way. According to the researchers, a fair punishment is unpleasant but offers relief: facing the consequence closes the episode and lifts the burden of guilt ^[10]. Escaping punishment produced the opposite arc, with initial relief giving way to guilt and sadness, particularly once the child learned a peer had been punished for the same thing ^[11]. Knowledge was the hinge. When the story character did not know how a peer had been treated, children judged that character's feelings from their own punishment or lack of it; once the character learned of the difference, children attributed sadness, anger and guilt regardless of what had actually happened to them ^[9].

For anyone writing a behaviour policy, the operational reading is that escalating severity buys less than most codes assume, and that the first enforcement decision of a term is the expensive one ^[3][4]. The cost is not the sanction. It is the visible asymmetry.

Two limits. This is vignette work: children judged picture stories, not the adults who actually govern them ^[1]. And the reported account gives combined figures rather than separate results for the Polish and American samples, so cross-cultural stability is not something it establishes ^[13]. Worth watching whether an explained exception - an accommodation, a documented reason - reads differently to a 7-year-old than an unexplained one, because that is the case real classrooms produce most often, and the reported design does not settle it ^[1][13].