Experiments in Majoritarian Bargaining

We investigate the predictive success of the Baron-Ferejohn model of legislative bargaining in laboratory environments. In particular, we use a finite period version of the bargaining game under weighted majority rule where a fixed payoff is divided between three players. We find that our subjects’ behavior is not predicted well by the Baron-Ferejohn model. The model predicts hardly better than a coin flip which coalition partner is selected by the chosen proposer, and proposers allocate more money to other players than predicted. A sizable number of proposals are rejected in the first proposal periods, and subjects who vote to reject a proposal on average receive a higher payoff from the new proposal. We find that a simple equal sharing rule yields point predictions that can account for 2/1 to 4/3 of all accepted proposals.