Improving Coordination in Assembly Projects

Abstract

We study assembly supply chains with multiple contractors. In such settings, the original equipment manufacturer (OEM) sometimes outsources component production to different contractors, and the final product revenue may depend on the performance of the slowest (or most unlucky) contractor. To mitigate this risk, the OEM may use a risk sharing contract, under which the OEM pays contractors after the project has been completed and the OEM has collected the revenue. Such risk sharing contracts are common in assembly supply chains, but they give rise to a coordination problem among the contractors because there exist multiple Pareto-ranked equilibria. This game has a similar structure to the coordination game, extensively studied in behavioral economics, and like in the coordination game, the equilibrium that emerges in the laboratory is the least efficient one, known as the risk dominant equilibrium. To mitigate this coordination failure, we introduce a new mechanism that involves informing contractors about the progress made by other contractors. We call this the information-feedback mechanism. We model the information feedback mechanism analytically, and find that for any number of contractors, there exists a minimum number of feedback periods that is required to achieve coordination. This minimum number of feedback periods increases with the number of contractors. We test our new mechanism in the laboratory with human subjects who have been incentivized based on their performance and found qualitative support for the model. Coordination improves with the number of feedback periods, although more gradually than the theory predicts. Coordination also deteriorates as the number of contractors increase, confirming the intuition that larger assembly systems are more difficult to coordinate.

Keywords: Behavioral Operations Management, Equilibrium Selection, Coordination, Information Feedback, Non-Cooperative Game Theory