Beyond the dividing pie: Multi-issue bargaining in the laboratory

Abstract

We design a laboratory experiment to study bargaining when parties need to agree on multiple issues. We find that bundling - the ability to make price offers on combinations of issues rather than separately - is critical for reaching agreement. We also find that giving bargainers access to more information about each other's valuations and costs does not raise efficiency, because the boost in agreement rates in small-surplus negotiations is offset by increased risk-taking and conflicting fairness preferences in large-surplus negotiations. Finally, we show that successful negotiations are characterized by an alternating offer structure, which emerges endogenously. It involves offers that split the difference between the two most recent demands, and it displays a higher probability of agreement vis-a-vis other formats of bargaining observed in our data.