認罪認罰“從寬”裁量模式實證研究

由於認罪認罰與其它量刑情節之間存在重合，一種觀點認為，認罪認罰的“從寬”適用需要依附於自首、坦白、積極賠償等實體性的從寬情形。另一種觀點認為，認罪認罰的“從寬”適用具有獨立性和專屬性，應當在其它量刑情節之外適用額外的從寬。通過採集自2016年認罪認罰從寬試點以來六個城市法院裁判的30129份醉酒型危險駕駛罪的判決書進行實證研究發現：首先，司法中對認罪認罰從寬的適用部分依附於被告人的犯罪情節、自首、坦白以及對被害人的“積極賠償”行為，但是相較於前兩個影響因素，“積極賠償”這一情節對於從寬適用的影響較弱。其次，在運用傾向得分匹配方法控制了相關情節之後，認罪認罰的被告人僅在自由刑的裁量上受到了較低幅度的從輕處罰，在罰金刑和非強制措施的適用上均沒有獲得顯著和普遍意義的從寬。最後，各個城市在從寬的範圍、模式和幅度上均具有懸殊的差距。

Due to the overlap between ''admission of guilty and acceptance of punishment'' and other sentencing factors, in previous literature, some argue that leniency should depend on other sentencing factors such as turn himself in, confession, and victim compensation. However, others hold that the leniency should be imposed independently and specifically. Therefore, defendants who admit guilt and accept punishment should receive extra leniency. This study collects 30129 DUI cases in six cities after the ''imposing leniency on admission of guilty and acceptance of punishment'' program was carried out in 2016. Empirical results show following findings: First, the application of the ''imposing leniency on admission of guilty and acceptance of punishment'' partially depends on the defendant’s crime severity, whether the defendant turns himself in or confesses, and whether the defendant compensates the victim. However, compared to the first two factors, the ''victim compensation'' factor has a relatively weaker effect. Second, using PSM to control for confounding variables, the results show that defendants who plead guilty only receive small leniency regarding sentencing outcomes. By contrast, the decision of pleading guilty has no significant effect on the fine amount and pretrial detention. Finally, there are significant regional disparities in terms of the scope, mode, and the magnitude of leniency imposed on defendants who plead guilty.