September 25, 2017 – A new paper by Dr. Alec Smith (together with coauthors from Caltech and Wharton) studies private information in dynamic, unstructured bargaining. The authors show how game theory can be used to make predictions about bargaining outcomes, even in complex dynamic bargaining environments. Dr. Smith and coauthors study this benchmark model in an experiment, and find that it can account for some but not all of the behavior. They use machine learning to demonstrate that the process of bargaining – how the bargainers interact – also influences whether a deal is reached. The work has implications for understanding deal making in a broad range of settings, from labor negotiations to climate change. The paper is forthcoming in Management Science.