Supply Chain Engineering

Principal Investigators: Ulrich Thonemann, Elena Katok

Most research on supply chain management assumes that the objective of the decision makers is to maximize expected profits. However, laboratory and preliminary field data suggest that order quantifies typically deviate from expected profit maximizing levels. The behavioral reasons for these deviations are not well understood. To address this issue we will focus on two subprojects: contract design and social behavior. In the first subproject, we will build and test hypotheses on potential objectives and rationales of individual decision makers, taking into account elements such as risk- and loss-aversion, over-confidence. Prospect Theory and decision making heuristics. To examine these, we will look at a setting where a single decision maker is offered a supply contract in a stochastic demand environment and decides on the order quantity. Our goal will be to identify the objectives and rationales of human decision makers in supply chains and build analytical models that predict how they respond to different supply contracts. Then, we will use these models for designing supply contracts that do not solely rely on the profit maximizing assumption. While the first subproject will focus on an in-depth analysis of individual decision makers in a stochastic demand environment, the second subproject will investigate settings in which decision makers interact, taking social behavior into account. We will first analyze a deterministic demand setting and then, after results from the first subproject become available, proceed with the stochastic demand setting. In our analyses, we will build on eariier work that shows that channel inefficiency can be partly explained by fairness preferences and trust and will analyze a number of supply chain contracts. Our focus will be the effect of different bargaining protocols on contract performance and how those affect cooperation. In the project, we will use economic models, laboratory experiments and field data.