关键词: 邻避风险沟通, 私人信息源, 信任驱动的学习, 非贝叶斯劝说, 机制设计

Abstract:

The rapid development and transformation of the economy and society have led our country to gradually enter a high-risk society. With citizens’increasing awareness of risks and rights， the NIMBY （not in my backyard） effect has become a significant source of social stability risk. Managing the NIMBY risk chain， which involves transforming “actual risk” into “perceived risk”， is crucial for preventing the externalization of NIMBY syndrome into social stability risk. Based on the asymmetry of actual risk information， it aims to design the optimal risk disclosure mechanism for the administrator to address NIMBY syndrome. The residents’ private risk information sources， global network externalities of residents’ behaviors， and trust-driven learning rule are considered to model two typical communication scenarios in NIMBY events： strategic interactions among residents and risk disclosure from administrator to residents.In this way， a network communication and persuasion model is developed based on the Network Cheap talk and Bayesian persuasion theories. It is assumed that the administrator’s disclosure occurs after residents’ private information learning and strategic communication within community. Thus， a Network Cheap talk model is constructed to depict the strategic communication process among heterogeneous residents， deriving the perfect Bayesian equilibrium and analyzing the consensus on community risk beliefs under the equilibrium condition， as well as changes in consensus based on different private information realizations. Based on this， a public risk disclosure mechanism is designed for the administrator. Since residents’ private information serves as uncertain information for the administrator， the optimal mechanism must implement differentiated recommendations for residents based on different private signals while satisfying the incentive compatibility condition. Additionally， drawing from conservative Bayesianism， an affine distortion approach is employed to characterize resident’s trust-driven learning rule.The findings indicate that the optimal public risk disclosure mechanism is a deterministic interval threshold recommendation mechanism， providing incentive-compatible differentiated recommendations based on residents’ private information. The more optimistic the residents， the broader the recommended risk acceptance interval. Meanwhile， the negative intensity and number of private information sources determine whether and how administrator should disclose risk and the effectiveness of such disclosure. Additionally， residents’ trust positively influences the mechanism’s persuasive effect. However， the expected payoff of pessimistic residents （high private information environment） is a decreasing function of trust， leading to a non-increasing relationship between trust and social welfare.The optimal risk communication mechanism for administrator and offers specific suggestions are offered to enhance communication effects from private information sources is clarified and trust management. Also， the model developed in this paper contributes to the non-Bayesian persuasion theory under the persuasion scenario considering one-to-many， continuous state space， and strategical communication among receivers.

Key words: NIMBY risk communication, private information sources, trust-driven learning, non-Bayesian persuasion, mechanism design