Self-Enforcing Commitment: A Dual-Process Account of Promise-Keeping

Why do people keep promises when breaking them would go unpunished? The common explanations rely on reputation or repeated interaction, but these require the possibility of consequences. I propose that promise-keeping operates as a default response tendency. Breaking a promise requires deliberate override, which consumes executive control, reducing capacity for later tasks. Anticipation of this cost can make keeping a promise rational even when breaking it carries no external penalty. While other accounts focus on dual-process conflict primarily as a source of moral error, this paper treats the cognitive cost of override as a functional asset: an internal penalty that can stabilize cooperation. In a preregistered one-shot trust game with UK adults (N = 514), trustees could promise to return money to a partner and then keep or break that promise. The sole manipulation was the bonus for a subsequent working-memory task: $1 or $20. As predicted, the rate of promise-breaking was significantly lower when later cognitive stakes were high (OR = 0.62, p = .019), consistent with the claim that anticipating a temporary reduction in cognitive efficiency deters promise-breaking.