Optimal dynamic incentive mechanism design for construction and demolition waste recycling with Bayesian learning

Abstract

A lack of a steady supply of high-quality construction and demolition (C&D) waste hinders the development of the recycling industry, and information asymmetry makes it difficult for a recycling company to determine incentive strategies for a collector. To address the problem, a dynamic moral-hazard model is formulated to obtain the optimal incentive mechanism. Bayesian learning is then introduced to update the estimation of a collector's private information. The results show that the collector is always motivated to voluntarily sustain a high-quality supply of C&D waste in the optimal mechanism. Additionally, the private information is gradually revealed by learning, which is conducive to controlling incentive costs. However, the collector is likely to misreport their information for higher profits, which reduces the learning accuracy and increases the incentive cost. Numerical analyses suggest that in response to misreporting, the recycling company should partially ignore the provided false information and extend the duration of each period of the contract over time, which can improve the learning accuracy. Therefore, this study helps to realize better recycling management from the perspective of interactive decision-making between the recycling company and the collector.