Convicting Corrupt Officials: Evidence from Randomly Assigned Cases
利用菲律宾法院案件随机分配给严格程度不同的法官这一自然实验，研究发现定罪腐败官员能改善地方公共财政管理并减少腐败，但效果主要体现在官僚而非政客身上。
Abstract Can the judiciary help root out government corruption? This article exploits the random assignment of court cases to justices who exhibit varying degrees of strictness to examine how convicting corrupt officials affects local government outcomes in the Philippines. I document that convictions improve the management of local public finances and reduce associated corruption. An exploration of mechanisms suggests that legal deterrence effects contribute to these findings. The results further indicate that convictions are effective at stifling corruption among bureaucrats, but not politicians. Consistent with this heterogeneity, convictions reduce electoral competition and consequently weaken accountability for incumbent politicians.