Learning by Doing in the Newsvendor Problem: A Laboratory Investigation of the Role of Experience and Feedback

We investigate learning-by-doing in the newsvendor inventory problem. An earlier study observed that decision makers tend to anchor orders away from the expected profit maximizing order. We study enhancements to the learning environment suggested by the behavioral literature. Extended experience improves performance some but feedback enhancements do little. A manipulation with a strong positive effect constrains inventory policy to a standing order for multiple periods, precluding responses to short term fluctuations in demand. With experience, our newsvendors collectively achieve over 90% of expected profit potential. It turns out that this kind of hands-on learning is more effective than providing newsvendors with an upfront statistical analysis that includes the expected profitability of orders. The results speak to the kinds of training and decision support that might improve the performance of inventory decision makers.