| Summary: | 碩士 === 臺灣大學 === 政治學研究所 === 98 === For long, court judges’ discretion in sentencing has always been questioned. This study thus attempted to open the black box of judges’ sentencing to examine if any inconsistency exists in their sentencing standards and causes of the inconsistency. The research sample consisted of a total of 1,591 verdicts for homicide cases made by district courts around the nation during 2003~2008. Based on factors affecting judges’ sentencing, four variables including defendant’s individual factor, criminal behavior factor, victim factor, and judge factor were selected and used for analysis. The data were analyzed using descriptive statistics, analysis of variances (ANOVA), and multiple classification analysis (MCA).
The main conclusions are as follows: (1) “Legal aggravation of punishment” is completely not influential: The stipulation about imposing aggravate punishment on repeated offenders of crime or offenders in cases where victims are minors has not been adequately reflected on judges’ sentencing; (2) Only a few non-legally prescribed sentencing factors affect judges’ sentencing decisions but not to a significant extent: This study listed 12 non-legally prescribed sentencing factors. Among them, only defendant’s age, gender, and the court affect judges’ sentencing decisions but not to a significant extent. Their explanatory power was also not high; (3) The legally prescribed sentencing factors (Article 57 of Criminal Code), such as criminal motivation, criminal method, criminal profile, and victim’s attitude, considerably influence judges’ sentencing decisions; (4) The most influential factor is “legal alleviation of punishment”. Many judges are used to slashing the original punishment by half first if legal alleviation of punishment is applicable to the defendant and then roughly adding a few months to the reduced punishment, without prudent consideration of the severity of the crime and a suitable proportion of punishment to be reduced. The coefficient of determination for this conclusion (R squared) reached 0.496. In social science, this value suggests high explanatory power.
Based on the above conclusions, this study proposed short, mid, and long-term goals for improvement of judges’ sentencing: (1) Short-term goal: For stipulations about aggravation and alleviation of punishment, different standards should be set up for different criminal conditions; (2) Mid-term goal: A sentencing standard for homicide should be set up; (3) Long-term goal: A special institution responsible for collecting data and empirical research of sentencing should be established.
|
|---|
