Do people stand by their commitments? Evidence from a classroom experiment

We test the fulfilment of commitments in terms of chosen effort that students intend to exert under an incentive contract in the absence of strategic interactions. We designed classroom experiments whereby students had to decide on whether to commit to their ex ante level of effort despite that they have no interest to do so ex post. The experiment consisted of three within-subject treatments that define different institutional contexts regarding information disclosure, evoking different levels of social visibility and commitment. In contrast to the theoretical predictions, a significant percentage of students tend to stand by their ex ante commitments regardless of the manipulation of the social context. Our findings corroborate the existence of a freezing effect by showing that promise keeping is sensitive to our manipulation of social visibility and self-commitment, with a stronger effect for public disclosure of the average level of effort than for public self-announcement of private choices. Our data also allows us to shed light on promise making behaviour. Contrary to promise keeping, there are order effects. However, the number of observations is too small to identify a difference between the baseline treatment and the one implying public self-announcement of level of effort.