Threats and demonstrations of power: experimental results on bilateral bargaining

We test the empirical effectiveness of threats in equilibrating bargaining power in simple bilateral bargaining games. Our experimental design is based on the two-player versions of the multibidding game (Pérez-Castrillo and Wettstein, 2001) and the bid-and-propose game (Navarro and Perea, 2005) that build on the ultimatum game and balance parties' bargaining power by auctioning the role of the proposer in the first stage. We show that, while both mechanisms implement the fair split in their subgame-perfect Nash-equilibrium, they induce different results in the laboratory. Subjects do not react to threats that lie off the the equilibrium path in the expected way, even if these threats are theoretically credible as they belong to the subgame-perfect Nash equilibrium. In particular, it seems that subjects feel the need to show their bargaining power as if punishment that never happens could not constitute a credible threat.