History is abundant with authoritarian rulers who reluctantly delegated governance powers to other institutions. Nevertheless, the intriguing question is why would democratic powerful rulers delegate powers voluntarily to (probably) contrarian institutions and tolerate activism that impedes the implementation of their election commitments. I develop a principal-agent model based on Fiorina’s (Congress: the Keystone of the Washington Establishment, Yale University Press, UK, 1977; Public Choice 39:33–66, 1982) blame-deflection hypothesis that politicians use the separation of powers principle as a shield against public criticism and accountability and addresses Stephenson’s (Journal of Legal Studies 32:59–89. https://doi.org/10.1086/342038, 2003) criticism. I show that in subgame-perfect equilibrium, the agent manipulates the principal. The extended model embeds experience and impartial judiciary and shows that the equilibria are preserved, but social welfare is lower.

历史上充斥着极权统治者，他们不情愿地将治理权下放给其他机构。然而，有趣的问题是，为什么民主强大的统治者会自愿下放权力（可能）反对机构并容忍阻碍其选举承诺实施的激进主义。我基于 Fiorina 的（国会：华盛顿机构的基石，耶鲁大学出版社，英国，1977 年；公共选择 39：33-66，1982 年）开发了一个委托代理模型，即政治家使用三权分立原则作为抵御公众批评和问责的盾牌，并解决斯蒂芬森的（法律研究杂志 32:59-89. https://doi.org/10.1086/342038, 2003）批评。我证明了在子博弈完美均衡中，代理人操纵委托人。扩展模型嵌入了经验和公正的司法，并表明保持了均衡，但社会福利较低。