論被告品格證據在國民法官審判中的證據能力及職業法官的指示：與香港、加拿大和英格蘭比較研究

Abstract

本篇論文討論在國民法官制度下，被告良好和不良品格的證據能力，以及其作為證據之後，職業法官需要發給國民法官的指示。

在《國民法官法》下，刑事審判引入六位非法律專業的國民法官，會同三位職業法官一同審理。但證據能力和證據調查之必要，交由職業法官在準備程序處理。《國民法官施行細則》第121條對被告品格證據作出規範，指其「不得作為證明本案犯罪事實成立與否之證據。但非以其品格推論本案犯行之成立，或有其他具體情事認與本案相關且必要者，不在此限」。唯不論該《施行細則》的本文還是其「逐條說明」，均沒有提出原則性、統一的標準，指導職業法官如何認定被告品格證據的證據能力。此外，根據該《施行細則》第161條，法院需要妥適審酌證據之證據能力及調查必要性，並「得權衡其對國民法官法庭以公正、客觀、中立方式審理之危害程度是否顯然高於其正面效益，妥適決定是否准許調查」。但具體需要考慮什麼因素，並無詳細說明。同時，若被告品格成為證據後，職業法官需要對國民法官作出什麼指示，為他們使用該等證據時設限，《法官對國民法官之指示參考手冊》亦沒有充分的指導。

本文首先總結台灣國民法官制度下對被告品格證據規範的含糊之處，然後比較香港、加拿大和英格蘭的有關法律，包括如何認定被告品格能否作為證據，以及法官對陪審團的指引，並從中吸收可取之處，摒棄不可取的規定。而本文核心的倡議為修改前述第121條，採納「原則性的法則」，規定被告品格證據須通過相關性測試，以及第161條的權衡，方有證據能力及准予調查。本文認為被告的品格可以作為證據證明其犯罪傾向及陳述的可信度，但基於該證據帶來的偏見風險，特別舉出不同的情況，說明其證據價值，並列出衡量證據價值和偏見需要考慮的因素，為職業法官作出較完善的指引。本文認為職業法官對證據能力的判斷，是控制被告品格證據偏見的第一個安全網。第二個安全網，是職業法官對國民法官的指示：當被告品格正式成為證據後，盡量防止國民法官誤用。本文為十種情況撰寫了供參考的指示，該等指示亦特別針對被告品格證據帶來的偏見，作出限制，進一步減低證據誤用的風險。

This thesis discusses the admissibility of a defendant’s good and bad character in citizen judge trials, and once such evidence is admitted, the directions needed to be given by professional judges to citizen judges.

Under the Law on Citizen Judges, six non-professional citizen judges will try a criminal case together with three professional judges. However, the admissibility of evidence and the necessity of inquiry are to be determined by a professional judge at preparatory proceedings. Section 121 of the Implementation Rules of the Law on Citizen Judges regulate a defendant's character evidence, which says, it “shall not be evidence proving whether the criminal facts are established or not, provided that it is not used to infer criminal acts from character, or there are other concrete matters recognised to be relevant and necessary.” Nevertheless, neither the Implementation Rules or their “Section-by-section Illustration” provide any principled and uniform standard, guiding professional judges how to determine whether a defendant's character evidence is admissible. In addition, according to section 161 of the Implementation Rules, the court shall properly examine the admissibility of evidence and necessity of inquiry, and “may consider whether its harm to the ability of a citizen judge court to try the case fairly, objectively and neutrally, substantially outweighs its positive effects, and properly determine whether to allow inquiry or not”. Yet there is no detailed illustration what factors are to be considered. Meanwhile, the Model Citizen Judges Instructions have not provided sufficient guidance to the situation that once a defendant's character evidence is admitted, what instructions professional judges need to give to citizen judges, limiting their use on such evidence.

This thesis first summarises the ambiguities of the law on a defendant’s character evidence under the citizen judge system in Taiwan. It then compares the relevant law in Hong Kong, Canada and England, including the determination of admissibility and the judges’ instructions to juries. It adopts the commendable parts and abandons the otherwise. The core argument of this thesis is to amend the aforementioned sections 121 and 161 and to adopt a principled rule such that a defendant's character evidence must pass the test of relevance and the balancing exercise under section 161 in order to be admissible and allowed to be inquired. This thesis suggests that a defendant's character may be evidence proving his criminal propensity and credibility. Yet, in the light of the risk of prejudice brought by such evidence, this thesis specifically explains its probative value in different scenarios, and lists out the factors needed to be considered when balancing its probative value and prejudice, providing better guidance to professional judges. This thesis suggests that the judgment as to the admissibility by professional judges is the first safety net to control the prejudice of a defendant's character evidence. The second safety net is the instructions given by professional judges to citizen judges. When a defendant's character evidence is admitted, the instructions aim to prevent citizen judges from misusing such evidence. This thesis provides model instructions for ten scenarios. They specifically limit the prejudice brought by a defendant's character evidence, further reducing the risk of misuse.