The Impacts of Prediction Technologies on Relational Contracts

The recent advances in data science and artificial intelligence have led to better and cheaper prediction machines, which improve business decisions in virtually all industries. We show that these advances, however, can do harm to the economic efficiency of supply chains that rely on relational contracts. This is because better predictions undermine the willingness of parties to commit to and trust each other long term. Using a model of relational contracts between suppliers and retailers who have different prediction abilities on market shocks, we identify and analyze an inherent trade-off between the new prediction technologies and the old trust-building mechanism in relational contracts.