Inventory decisions made at a centralized level often rely on demand forecast information passed from regional managers within a supply chain. Such managers often have unique insights into the demand patterns at their local sites that can help inform how much inventory to order for the system as a whole. Problems can arise with this setup, however, if these managers have incentives to misreport their forecasts and the central planner (CP), in turn, mistrusts this information. The goal of our research is to shed light on the existence, magnitude, and causes of such coordination problems using a combination of analytical models and behavioral experiments. The analytical analysis reveals that incentives are misaligned in this setting and no truth-telling equilibrium exists in general, unless inventory competition or demand uncertainty is removed. With this theoretical grounding, we conduct a series of controlled laboratory experiments to test the magnitude of these problems and how they are impacted by inventory competition, strategic concerns regarding the other parties, and the demand environment. We find that inventory competition and market uncertainty harm the efficacy of forecast sharing and channel efficiency, while the possibility that untrustworthy reports are detected (identical local demand information) does not deter regional managers from misreporting their forecasts. Removing strategic concerns regarding the CP (automating the central ordering decision) does not improve profit but reduces order variability. Information sharing appears to be a dominant policy as it significantly improves profit in all treatments compared with no information sharing.