This grant provides funding for determining the role of trust in forecast information sharing. Consider, for example, the process through which a manufacturer (such as Hewlett-Packard) solicits forecast information from a retailer (such as BestBuy). The manufacturer uses this forecast information to plan for production and capacity. When the communication is non-binding, the retailer has an incentive to inflate demand forecast to assure abundant supply. The lack of credible information sharing leads to well-documented catastrophic outcomes. Yet, some firms manage to share information credibly even when communication is non-binding. In this research, predictive mathematical models and human-subject experiments will be developed to determine how the level of trust and trustworthiness among firms affect forecast information sharing and decision-making. These models will be validated with empirical data to ensure their applicability and abilities in predicting actions.
The primary goal is to develop a scientific understanding of trust and trustworthiness in strategic forecast information sharing. A new mathematical theory will be developed to model trust and enhance the neoclassical economic models (such as, game-theory) that assume humans make decisions based on pecuniary payoffs with self-interested motives. The new theory will provide a framework to estimate the level of trust (that the manufacturer has for the supplier's information) and trustworthiness (that impedes the retailer to inflate forecast information). This new model will help prescribe policies to better coordinate forecast information sharing and optimize decisions, such as capacity investment. The research will determine when trust can facilitate cooperation and how trusting behavior alters a firm?s forecast management and contracting strategy.
Conceivably, this research may also help develop a new paradigm in designing contracts that include considerations of trust, non-pecuniary issues and human emotions. Since this project is in collaboration with Hewlett Packard, the results is likely to be transferred into business practices.
