政策认知分歧、干中学与层级政府治理

Cognitive Divergence in Policy,Learning by Doing and Hierarchical Government Governance

本文将贝叶斯劝说引入中国式财政分权的政府治理框架,构建上下级政府间存在认知分歧的两阶段博弈模型,从理论上解释政府"治理失灵"与政策有效性并存的矛盾现象。研究发现:政策实践结果形成了对上下两级政府的双向贝叶斯劝说,任务结果考核制度既是有效的激励手段,也是有效的控制手段;"干中学"机制使得财政分权的激励效应存在门槛条件,仅当财政分权程度高于一个阈值时,建立考核制度才有正面激励作用,否则反而会产生负面激励;上级政府相机抉择时,下级政府的政策执行具有协商和权衡的作用,上下级政府能够在政策选择与执行过程中形成"共识型"体制。本文为农地制度变迁、扶贫工作中的政府行为、考核制度优化等改革实践提供理论解释。

This paper constructs a two-stage game model within the governmental hierarchy with cognitive divergences in terms of policy by introducing Bayesian persuasion into Chinese-style fiscal decentralisation governance framework,which theoretically explains the puzzle of the contradictory coexistence of governance failures and policy effectiveness.The results of the study illustrate that the practice in policy implementation forms a two-way Bayesian persuasion in governmental hierarchies due to the dual role of incentive and control of the performance appraisal system.The"learning by doing"mechanism creates threshold conditions on the incentive effect of fiscal decentralisation.The establishment of a performance appraisal system only provides a positive incentive effect when the degree of fiscal decentralisation is above a certain threshold value.Otherwise,a negative incentive effect occurs.The policy implementation of a lower-level government plays a consulting and weighting role when a higher-level government makes a discretionary decision,able to form a"consensus type"mechanism during the policy selection and execution process.This paper also provides theoretical explanations for the implementation of the rural land system change reform,the government’s behaviour in poverty alleviation and the optimisation of the performance appraisal system.