Seminar Sidelights | Four Years of the Judges Act Amendments: Lecture 1 of the Pluralistic Judicial Accountability Series

The Judicial Reform Foundation has long been concerned with the oversight of incompetent judicial officers. To foster dialogue on shared expectations and improvement goals for judicial professional ethics across subjects, objects, and methods, we co-hosted the "Four Years of the Judges Act Amendments: Pluralistic Judicial Accountability" seminar series with the Judicial Reform Committee of the Taiwan Bar Association, exploring "oversight mechanisms for incompetent judicial officers" and "self-regulation and external regulation of attorneys."

This event was the first of the seminar series. The first half was moderated by Attorney Chiang Jung-hsiang, Chair of the Judicial Reform Committee of the Taiwan Bar Association. Li Hung-yi, Supervisory Investigator at the Control Yuan, and Attorney Li Ming-ju, Deputy CEO of our Foundation, delivered reports on the Control Yuan's and civil society groups' observations regarding the individual case evaluation system four years after the Judges Act amendments. Professor Chiang Shih-ming, Professor at the College of Law of National Chengchi University and former member of the Judges Evaluation Commission, and Attorney Chien Chien-jung, former Judge of the Supreme Court, were invited as respondents.

The second half was moderated by Attorney Huang Hsu-tian, Chairperson of our Foundation, with the Judicial Reform Committees of the Taichung, Tainan, and Kaohsiung Bar Associations sharing how bar associations participate in oversight as entities eligible to request evaluations. Finally, Attorney Lin Yung-sung, Managing Director of our Foundation, responded with observations from civil society.

Four-Year Observation Report on the Implementation of the New Judges Act Released: Li Ming-ju Criticizes It as "Old Wine in New Bottles"

During the seminar, the JRF released the "Civil Society Annual Observation Report on the Fourth Year of Implementation of the Amended Judges Act," pointing out three repetitive issues: individual requests from citizens remain difficult to substantiate; grounds for "not initiating" evaluations are mostly "requests for evaluation based on legal interpretations" and "evidently groundless"; informants filing non-initiating requests account for the majority; and the anonymity mechanism for resolution documents is virtually defunct.

In the fourth year of the new Judges Act implementation, both the Judges Evaluation Commission (JEC) and the Prosecutors Evaluation Commission (PEC) handled 11 evaluation cases in total, with the majority of resolutions being decisions not to initiate evaluations. Among these, citizen-requested cases resulting in a substantiation decision amounted to only 1 in the JEC and zero in the PEC. For the JEC, the most common reason for not initiating an evaluation was "evidently groundless," followed by "requests for evaluation based on legal interpretations"; for the PEC, the primary reason was "evidently groundless," followed by "the requester lacks standing." Further analysis revealed that "informants" constituted the overwhelming majority.

Because published resolution documents are de-identified—such as omitting case numbers—it is impossible to study repeatedly evaluated cases or review case details. However, media reports are quite detailed, creating a paradoxical situation where cases can be reviewed backward through the media.

Notably, the only citizen-requested case substantiated in four years, the "Judge Li Chao-jung Case," involved a judge presiding over a tort damages compensation case who urged the parties to jointly suspend the proceedings by agreement, stating that the clerk would notify them later. When the four months were up, the judge had been rotated and the clerk failed to remind them, resulting in the case being deemed withdrawn. Judge Li Chao-jung claimed in her defense that she had given similar directives in medical lawsuits, indicating that she had engaged in similar behavior prior to this case. Furthermore, Judge Li had already been disciplined before this case for poor courtroom demeanor, highlighting the importance of investigating a judge's personality and behavioral profile.

JRF Proposes Nine Recommendations:

1. Strengthen statements and investigative assistance: Help citizens overcome obstacles in submitting written briefs and clearly explain misconduct.
2. Provide the right to review audio recordings: Allow citizens to review courtroom recordings and express their opinions prior to resolution.
3. Enhance the investigative efficacy of the PEC: Review staffing allocations and investigative workflows to boost investigative proactivity.
4. Comprehensively review misconduct circumstances: Investigate whether the evaluated person has other related complaints or misconduct cases.
5. Lower barriers for disadvantaged citizens: Cooperate with legal aid organizations and establish local branches.
6. Expand evaluation request standing: Grant "litigation agents" and "informants" the right to request evaluations.
7. Publish annual implementation reports: Release annual reports on evaluation performance for public review and oversight.
8. Return to openness and transparency: Align the publication of resolution documents with the principles governing general judgment documents.
9. Comprehensively review management systems to prevent bad money from driving out good (enhance systematic methods for internal weeding out of incompetent personnel; optimize or redesign comprehensive evaluation systems).

Li Hung-yi Responds from a Practical Observational Perspective, Echoing JRF's Views

Synthesizing the content of resolution documents, the evaluation commissions' stance and standards are: evaluations must prioritize safeguarding judicial independence; evaluations must not impair judicial independence; legal interpretations do not constitute grounds for evaluation; legal interpretations are to be remedied through the appellate system; and the evaluation commission is not a fourth-instance court. Both the JEC and PEC emphasize judicial independence and the protection of the core of investigation.

Cases referred for evaluation by agencies and organizations have a high substantiation rate; conversely, under the new system, cases independently requested by citizens have a low substantiation rate. It is recommended to improve the search method for resolution documents of the judges and prosecutors evaluation commissions, which are currently decontextualized and difficult to search.

As seen in 113-Ping-Zi No. 5, the party in this case had already been subjected to job oversight pursuant to the Judges Act during a previous hearing due to poor demeanor. This highlights the importance of personality profiles during the evaluation process. The formation of a personality profile requires cross-referencing internal factors (comparing agency reputation and past cases from evaluation commissions) with external factors (opinions from civil society groups).

Returning to original intent, the purpose of evaluations and disciplinary actions (cited from the legislative intent of the Judges Act) is not to evaluate or censure a judge's individual misconduct acts one by one, but rather to comprehensively evaluate the personality manifested by all of a judge's misconduct acts in order to determine the responsibilities the judge should bear. When it is known that an evaluated judge has other circumstances warranting evaluation, those matters should be investigated and deliberated concurrently to achieve procedural economy and prompt evidence collection.

Simultaneous with accountability, there should be corresponding配套 measures. Judges and prosecutors possess an official character and basically cannot select their cases. Endlessly increasing obligations without corresponding配套 measures turns into a reason for mitigating responsibility and fails to meet public expectations.

Chiang Shih-ming: JRF Cannot Sit Idly By Regarding Judicial Infliction of Harm

In the current operation of the evaluation system, due to a lack of clarity regarding internal details, it is impossible to reconstruct the procedural flaws during case handling—lazy handling goes unaddressed, and cases being buried go unaddressed. When prosecutors encounter lazy and periquictory non-prosecution dispositions, it falls under prosecutorial discretion and goes unaddressed. For judges who fail to maintain the requisite legal intelligence in handling cases, even repetitive errors are rationalized as discretionary formation and judicial independence.

Myths within the existing system include the notion that legal interpretations cannot be evaluated. But what is a legal interpretation? Is it equivalent to a judgment? Acts of recognizing facts and applying law? Direction of proceedings? Direction of litigation? The core of adjudication?

Judicial independence and prosecutorial discretion are means by which a state governed by the rule of law enables two agencies to accomplish their missions, not ends in themselves. They should not become a fortress for judges or a Great Wall of responsibility. If means are mistaken for ends, many things cannot be addressed pragmatically.

Taking American standards as an example, if a California judge commits a legal error, it can be reviewed under exceptional circumstances—such as repetitive errors, alongside abuse of authority, contempt for basic rights, intentional and malicious willful illegality, and any purpose other than the faithful execution of judicial duties. These behaviors must be uncovered and halted.

To be frank, Taiwan currently experiences judicial infliction of harm, where people suffer secondary and tertiary victimization in the courtroom. How can the JRF sit idly by? If there is a severe violation of logic and empirical rules, and someone is a recidivist who cannot possibly be a recidivist, does this not qualify an incompetent judge? If none of this can be punished, it means the JRF has not tried hard enough.

Mandatory reasoning in judgments treats people as human beings rather than treating judges as gods. Only temple keepers and altars can proclaim divine will without providing reasons. Judicial reform must also address matters objectively. We must evaluate lazy judges; judges who previously ignored fully-delivered stocks while pushing aside or burying cases have instead become prophets. We will continue to issue strong condemnations.

Chien Chien-jung: Removing the JRF Causes Evaluations to Wither

Thirty years ago, corrupt judges were everywhere, and evaluations easily found them. Today, there are almost no corrupt judges. The new-generation evaluation of 2025 should be a battle of judicial pathways and values: what kind of judiciary do we want to give the people—one that protects human rights or one that protects authoritarianism? Is human rights paramount, or is preserving the existing legal system paramount?

Currently, the channels for seeking evaluation are very unfriendly to the public, and the most knowledgeable and capable civil society organization has been removed by barring it from acting as an applicant. Removing the JRF causes evaluations to wither, serving the objectives of high-ranking Judicial Yuan officials. We should adopt a dual-prosecutor system.

At present, the Judicial Yuan itself refers the most cases to the evaluation commission, but in reality, the Judicial Yuan could simply refer them to the Control Yuan or the Disciplinary Court. The Judicial Yuan itself lacks the courage to decide to refer cases to the Personnel Review Committee, making the Personnel Review Committee the most important driving organ. This matter of referral organs needs to be reviewed.

I support evaluating based on the "unity of misconduct behavior" rather than looking solely at individual cases when in reality the person is a habitual offender. The terminology used in evaluations is problematic; the word "individual" should be removed, otherwise it falls into the concept of no petition, no action.

I am also confused by the rule that legal interpretations cannot be evaluated. Taking theft with a lethal weapon as an example, for Judge Chien, this would never be constituted, but other judges would find it constituted. The problem lies with those other judges, because neither parties nor lawyers would bring such an interpretation for evaluation, yet no three judges have the courage to refer it for evaluation. Following a legal interpretation simply because a precedent says so, without thinking, represents a lack of cognitive ability.

Experience Sharing by Attorney Huang Sheng-fen of the Taichung Bar Association

The Taichung Bar Association distributed evaluation materials to all lawyer members in 2011, 2014, 2015, and 2016, but response rates were very low. Following the amendment of the Judges Act, a total of 1,624 copies were distributed in 2023, with 652 returned, yielding a response rate of 40.1%.

The evaluation subjects included judges of the Taichung High Administrative Court, the Taichung Branch of the Taiwan High Court, the Taichung District Court, and investigation/public prosecution prosecutors of the Taichung District Prosecutors Office. The results were analyzed by the Feng Chia University Quality Assurance Center.

Evaluation items varied based on whether the subject was a judge or an investigation/public prosecution prosecutor. For judges, the evaluation items were questioning demeanor, conduct of litigation, quality of judgments, and moral integrity. For prosecutors, the items were interrogation/court-attendance demeanor, investigative procedural actions/court-attendance performance, quality of written documents, and moral integrity. Evaluation grades were categorized as 5: very good, 4: good, 3: average, 2: bad, and 1: very bad.

The results showed that whether judges were from the Taichung High Administrative Court, the Taichung Branch of the Taiwan High Court, or the Taichung District Court, their average scores were all 4 or higher, while the District Prosecutors Office scored around 3.8. No judges from the Taichung Administrative Court scored below 4; 14 judges from the Taichung High Court Taichung Branch scored below 4; 5 judges from the Taichung District Court scored below 3.5, concentrated in the civil division; and 6 prosecutors from the Taichung District Prosecutors Office scored below 3.5.

This demonstrates that the performance of most evaluated individuals is affirmed. The Taichung Bar Association only publicly announces the top five best-performing individuals in each agency; those with lower scores indeed exhibited certain performance issues, proving that poorly performing judges and prosecutors can be highlighted through evaluation methods, especially when the sample size is sufficient. Notably, some lower-scoring individuals also held positions as chief judges or presiding judges, indicating that courts decouple evaluations when selecting chief or presiding judges.

Experience Sharing by Attorney Cheng Chia-hui of the Tainan Bar Association

The Tainan Bar Association has been conducting judicial officer evaluations since 1995. At that time, paper forms were used, listing all judges and prosecutors in a very thick volume. In 2009, the Taiwan Bar Association proposed a unified evaluation form. Later, in response to the digital age—since paper evaluations were only conducted once a year—an online evaluation system was fully adopted in 2021, and paper forms were no longer issued, allowing lawyers to fill them out immediately after finishing a court session. However, the number of participants dropped the second year, so this year both paper and online methods are being used concurrently.

To encourage lawyers to participate more, the incentive bonus, originally distributed once a year, was changed to twice a year (in the first and second halves). Yet, it was found that the participants remain the same people, which is the current dilemma encountered.

Evaluation results are not published comprehensively; only the top-performing individuals are announced on the website. Lower scores are not published, but evaluation reports are brought along when paying courtesy visits to courts and prosecutors offices.

Experience Sharing by Attorney Liang Chih-wei of the Kaohsiung Bar Association

The Kaohsiung Bar Association conducted evaluations in 2017, with results released in 2019. Following the Judges Act amendments, it currently seeks to drive a transition and launch an app, which is in the testing phase. We hope the Taiwan Bar Association will take the lead in promoting digital evaluations.

Experience Sharing by Attorney Lin Yung-sung, Managing Director of the Judicial Reform Foundation

The Judicial Yuan promulgated the Regulations on Judicial Evaluation in 1996, which were abolished in 2012 due to the implementation of the Judges Act. In 1998, the JRF conducted a comprehensive evaluation of 76 criminal court judges at the High Court and 47 criminal court judges at the Taipei District Court, grading them on three criteria: questioning demeanor, quality of judgments, and moral integrity. After statistical calibration, the names of the judges were published. Instead of feeling ashamed and apologizing, the judges held a press conference to launch attacks and filed a private prosecution for defamation against the JRF chairperson and executive director, which ultimately resulted in a final acquittal.

The purpose of comprehensive judicial performance evaluations in 20 U.S. states is to provide judges with self-improvement and align public expectations with the judiciary. Evaluation items include legal proficiency, integrity, impartiality, communication skills, professionalism, demeanor, and administrative capability. Methods include interviews, public comments, confidential lawyer reviews, judicial performance statistics, judicial judgments, questionnaires, courtroom observations, and disciplinary records. Committees are composed of lawyers and non-legal professionals. Every judge is evaluated once every three years, with results categorized as favorable or unfavorable, and some states make them public to the masses.

Currently in Taiwan, public trust in the judiciary is low. Although people can directly request individual case evaluations, not a single case has been referred to the Disciplinary Court. Performance appraisals lack objective and fair assessment, the quality gap in judicial service performance is too wide, and the current reality lacks a weeding-out mechanism for incompetent judges. Taiwan indeed needs a comprehensive judicial evaluation or review system.

It is recommended that an independent evaluation committee or foundation be established to enhance trust and participation among evaluators. Comprehensive judicial evaluation is not a disciplinary measure and should not be directly linked to elimination. However, if a judge ranks in the bottom 10% (or 20%), they should be evaluated annually with increased depth; if individual case evaluation grounds are discovered, the matter should be referred to the individual case judges evaluation committee.

Judges ranking in the top 10% (or 20%) of comprehensive judicial evaluations may have their names published and receive awards, whereas those ranking in the bottom 10% (or 20%) whose re-evaluations reveal individual case evaluation grounds should also have their names published.