主题：Limited Foresight Reasoning in an Infinite-Horizon Supply Chain Bargaining Game（无限期讨价还价供应链成员的有限推理能力行为分析）

主讲人：李娟  南京大学工程管理学院

时间：2023年10月7日15:00

地点：学院A421

摘要：

Abstract: This paper proposes a limited foresight-reasoning model to explain the experimental observations in an infinite-horizon supply chain bargaining game in which two supply chain parties, a supplier and a retailer, alternately offer contract proposals and decide whether to accept or reject. The proposed model shows that the hierarchy of level-𝑘 reasoning implies a hierarchy of limited foresight and converges to the perfect equilibrium as the reasoning level goes to infinity. To validate this model, we check the explanatory power of our model in an infinite-horizon supply chain bargaining game with discounting. Our experimental results show that when the players’ discount factors are relatively low, the model of lower-level foresight reasoning explains a substantial proportion of observations. That is, the players tend to exhibit a lower foresight reasoning level when their or their opponent’s discount factor is lower. The fairness motivation for equal split also plays a vital role in all treatments. We further propose structural estimation models to evaluate the effects of different motivations. We find that supply chain parties’ contract proposals are influenced by their intention of limited foresight reasoning and the perception of fairness motivation. The relative importance of these two factors is regulated by the magnitudes of and differences between the bargainers’ discount factors. Our experiments and simulations reveal that in treatments wherein the supplier holds a significant advantage over the retailer in terms of the discount factor and being the first mover in the first bargaining round, most suppliers exhibit lower levels of foresight reasoning, and the strategy based on limited foresight reasoning can yield a higher supplier payoff than that predicted by the equal split or perfect equilibrium.

摘要：当供应商和零售商之间进行基于轮流出价的无限期讨价还价时，他们会表现出有限推理行为。本文采用Level-k模型刻画供应链成员的有限推理行为。实验数据表明，当供应链成员的或其对手的贴现因子较低，其推理层级水平也较低。供应链成员也表现出了公平偏好倾向。本文采用结构模型估计了供应链成员的有限推理层级和公平偏好，发现，这两个行为因素带来的影响受供应链成员贴现因子相对大小的影响。采用实验数据和数值模拟分析方法发现，相对于具有公平偏好的博弈者，具有有限推理行为的博弈这能够从其行为特征中获得较多收益。

Co-authors: Jun Feng, Yulan Wang

合作者：冯骏，汪玉兰

主讲人简介：

李娟，南京大学工程管理学院教授，博导。主要研究领域为行为运营管理、供应链与物流管理。她的研究工作主要发表于Decision Science, European Journal of Operations Research, International Journal of Production Economics, International Journal of Production Research，管理科学学报等学术期刊。她指导的学生的研究工作分别获得2019年和2020行为运筹学与行为运营管理研讨会优秀学生论文。她还出版了4本与学术研究相关的普及读物，其中一本获得江苏省第15届哲学社会科学优秀成果奖（普及成果奖）。她主持国家自然科学基金项目4项，被评为《系统管理学报》2021-2022年度优秀审稿人，现任中国运筹学会行为运筹与管理分会理事。