Is earned bargaining power more fully exploited?

In past experiments involving a wide range of bargaining settings, individuals exploited their bargaining position less fully than standard theory predicts. Typically, these experiments allocated bargaining position randomly, so that bargainers, viewing their position as unearned, may have been reluctant to fully exploit it. We investigate the impact of earned bargaining power using theory and experiment. In our “Earned” treatment, disagreement payoffs – and hence bargaining power – are based on performance on a real–effort task. In our “Assigned” treatment, subjects perform the task but disagreement payoffs are randomly assigned. Our “Notask” treatment is like the Assigned treatment but without the task. Comparison of our Earned and Assigned treatments provides our main result: subjects are more responsive to changes in bargaining position when it is earned. Responsiveness is also often higher in our Assigned treatment than in our Notask treatment – suggesting a possible effect of merely including a status–irrelevant task – though these differences are usually insignificant.