Debiasing budget judgment: experimental evidence from politicians and civil servants

Purpose Public budgeting requires judgment under uncertainty, yet evidence on whether debiasing interventions improve such judgment remains limited. Although research has documented bias effects in public decision-making, much less is known about whether simple debiasing interventions work in budgeting contexts and whether their effects depend on the type of judgment and the way interventions are embedded in decision contexts. Design/methodology/approach The article reports a preregistered survey experiment among 1,230 politicians and civil servants. It compares three debiasing interventions (generic, realistic and innovative) across two judgment phenomena relevant to public budgeting: anchoring and loss framing. Treatment effects are analyzed quantitatively, supplemented by a descriptive analysis of written justifications in the loss-framing experiment. Findings Anchoring effects are strong and persist across all conditions, with no intervention reducing their influence. In the loss-framing experiment, no overall framing effect is observed. Only the realistic intervention is associated with lower allocations relative to the loss-framed condition. Descriptive evidence suggests that interventions are associated with different patterns in how participants justify their decisions. Originality/value The article shows that debiasing effects in public budgeting are conditional rather than generalizable. More specifically, it demonstrates that effectiveness depends on how interventions are embedded in decision contexts, supporting an ecological perspective on budget judgment.