Taiwan's 400-Point Bar Exam Threshold: From a Steep Drop in Passers to Zero Disqualifications in Four Years
In 2018, Taiwan’s Ministry of Examination introduced a new passing floor to the national bar exam: candidates had to score at least 400 points across the core legal subjects. In 2019, the second year under the rule, 355 candidates whose total scores placed them within the passing top 33% were disqualified for falling just short of the core-subject floor—driving the total number of new lawyers down to a low of 549.
Just two years later, when the results of the late-2021 bar exam were released, the number of passers rebounded sharply to 940. In its official press release, the Ministry of Examination announced that not a single candidate failed because of the 400-point threshold that year.
Since then, annual pass counts have held above 900, reaching a record high of 1,042 in 2025. The threshold remains on the books, yet it now disqualifies only about 10 to 21 candidates a year—almost nobody in practice.
A passing floor that once eliminated over a third of otherwise-qualifying candidates lost its bite within four years. The arc of this institutional experiment goes back to the two-stage bar exam reform of 2011.
The 2011 Two-Stage Reform and the Surge in New Lawyers
Before 2011 (ROC Year 100), Taiwan’s bar exam typically admitted 300 to 500 new lawyers a year. In 2011, the Ministry of Examination overhauled the system, replacing the single make-or-break sitting with a two-stage exam:
- Stage 1 (Objective Multiple-Choice Exam): Screens in the top 33% of test-takers who complete all sessions to advance to Stage 2.
- Stage 2 (Subjective Essay Exam): Written and graded by university law professors. It covers four core legal blocks—Constitutional and Administrative Law, Civil Law and Civil Procedure, Criminal Law and Criminal Procedure, and Commercial Law (800 points combined)—plus Chinese Composition (100 points). Starting in 2015 (ROC Year 104), a 100-point “elective subject” chosen from four tracks (Intellectual Property Law, Labor and Social Law, Tax and Finance Law, or Maritime and Ocean Law) was formally added, bringing the total possible score to 1,000 points.
At the time, passing was governed purely by relative ranking (norm-referenced grading): any candidate whose Stage 2 total score ranked in the top 33% of test-takers completing all sessions (or the top 33% within each elective track once electives took effect) earned a bar license. Because each stage took the top third, the compound pass rate (33% × 33%) settled at roughly 10% of all initial test-takers.
In 2011, the first year of the new system, the number of admitted lawyers surged past 900 to 963. From 2012 through 2017, annual pass counts remained steady at 800 to 900 (for example, 860 in 2016 and 924 in 2017). Producing nearly a thousand new lawyers a year, this period became known in Taiwan’s legal community as the “era of mass admission.”
Doubling the annual intake quickly triggered pushback from practicing attorneys. The Taiwan Bar Association (now the National Bar Association) and senior practitioners sounded the alarm: legal market demand had not doubled alongside the supply of new lawyers, and newly admitted attorneys were struggling to find law firms for their mandatory internship. Reports surfaced not only of underpaid traineeships, but of “pay-to-intern” arrangements and “vagrant lawyers” unable to secure training slots or steady practice. Some practitioners also complained that a minority of new entrants could barely draft a complaint or answer without glaring flaws.
Under pressure from the bar association and the legal profession, the Examination Yuan and the Ministry of Examination faced growing demands for a “headcount cap” to bring down the annual influx of new lawyers.
The 2017 Policy Bind: Using a 400-Point Floor as a Proxy for a Quota
Confronted with calls to cut admissions, the Ministry of Examination found itself caught in a structural bind: legally, the bar exam is classified as a “qualification exam” designed to verify baseline competence to practice. It cannot lock in a fixed headcount like a civil service recruitment exam, nor does it benefit from upstream enrollment caps like the national medical licensing exam.
Civil service exams and the judicial officer exam (for judges and prosecutors) are “recruitment exams,” where the government admits exactly as many top-ranking candidates as there are job openings. Meanwhile, although the medical licensing exam is a qualification exam with a pass rate consistently between 80% and 90%, the Ministry of Health and Welfare and the Ministry of Education strictly cap upstream university admissions at roughly 1,300 medical students per year, removing any need to restrict supply through the licensing exam.
In legal education, by contrast, the Ministry of Education imposes no quota on the proliferation of university law departments and graduate institutes, allowing thousands of law graduates to flood exam halls every year. Under the Professional and Technical Personnel Examinations Act and constitutional protections for the right to work and the right to take public examinations, the Ministry of Examination could not simply decree a flat cap of 500 new lawyers a year without explicit statutory authorization.
Unable to control university enrollment at the front end or slash the pass rate by fiat at the back end, the Ministry of Examination rolled out a compromise in August 2017. It amended the Regulations Governing the Senior Examination for Professional and Technical Personnel: Bar Examination, keeping the relative “top 33%” cutoff while layering on an absolute score floor (criterion-referenced grading).
Beginning in 2018, the new rule required that after excluding Chinese Composition (100 points) and the elective subject (100 points), a candidate’s combined score across the four core legal subjects (800 points possible) had to reach at least 400 points—half of the possible core points, or an average of 50%. Any candidate who scored below 400 on the core subjects failed automatically, even if strong marks in Chinese and the elective pushed their overall total into the top 33%.
NOTE
The Ministry of Examination’s official rationale was to “ensure that admitted candidates possess core legal competence and prevent candidates from passing merely by running up high scores in their elective subject and Chinese Composition.” In a tightly packed relative ranking where every point counts, gaining an extra 15 to 20 points on an elective could easily vault a candidate hundreds of places up the rankings, offsetting weak performances in the core subjects. Across the legal community, however, the reform was widely seen as “setting a passing standard in form while enforcing a headcount cap in substance”—a mechanism aimed at pushing annual admissions back down from the 800–900 range.
2018–2020: Unscaled Scores and Street Protests
Once the rule took effect in 2018 (ROC Year 107), the 400-point floor immediately compressed the number of passers and sparked the most intense wave of protests in the history of Taiwan’s legal exams over the next two years:
| Exam Year (ROC / CE) | Stage 2 Passers | Stage 2 Pass Rate | Disqualified by 400-Point Rule (Top 33% Overall, Core < 400) | Practical Impact of the Threshold |
|---|---|---|---|---|
| 105 (2016) | 860 | 9.87% | N/A (pre-threshold) | No threshold (baseline period) |
| 106 (2017) | 924 | 9.98% | N/A (pre-threshold) | No threshold (baseline period) |
| 107 (2018) | 759 | 8.58% | 107 | First year in effect; noticeable drop in admissions |
| 108 (2019) | 549 | 6.12% | 355 | Disqualifications peak; pass rate hits an all-time low |
| 109 (2020) | 650 | 6.76% | ~300 | Sustained squeeze; triggers mass litigation and protests |
| 110 (2021) | 940 | 10.32% | 0 | Threshold abruptly stops biting; zero candidates disqualified |
| 111 (2022) | 913 | 9.77% | 10 | Admissions return to the 900+ norm |
| 112 (2023) | 988 | ~10.0% | 21 | An eight-year high at the time |
| 113 (2024) | 980 | ~10.2% | 18 | Stabilizes near 1,000 new lawyers per year |
| 114 (2025) | 1,042 | ~9.9% | 15 | Record-high admissions; disqualifications remain marginal |
Note: the 2018–2020 figures are the official counts of candidates rejected specifically for scoring under 400 on the core subjects; from 2022 on, the Ministry publishes only a combined “rejected among full attendees” total (candidates completing all sessions), which includes zero-score and other mandatory-fail reasons.
In 2018, the rule’s debut year, 107 candidates were eliminated by the threshold, pulling the pass count down to 759. Without the floor, 866 candidates would have passed—right in line with previous years—but 107 people inside the top-33% cutoff were summarily cut.
In 2019 (ROC Year 108), tougher exam prompts combined with unadjusted grading standards caused 355 candidates in the top 33% to fall below the 400-point line. Total admissions plunged to 549, and the effective overall pass rate dropped from its historical ~10% norm to just 6.12%. In 2020 (ROC Year 109), roughly 300 candidates (294 to be exact) were similarly eliminated.
Why did a fixed 400-point cutoff produce such wild swings in annual pass counts?
On an objective multiple-choice test, the answer key is fixed and a score reflects an exact number of correct answers. Stage 2 of the bar exam, however, consists entirely of open-ended essay questions graded by university professors. There is no rigid answer key, and individual graders’ internal benchmarks shift from year to year. One year features esoteric prompts and stingy grading; the next brings mainstream questions and more generous marks. A raw score in one year does not carry the same weight as the same raw score in another.
Imposing a rigid 400-point cutoff across different cohorts without test equating to calibrate for exam difficulty and grading severity immediately exposed a fundamental flaw.
Comparing Taiwan’s bar exam with legal licensing exams in the United States and Japan highlights the psychometric gap:
| Dimension | Taiwan Bar Exam (Stage 2) | U.S. Uniform Bar Exam (UBE) | Japan’s National Bar Examination |
|---|---|---|---|
| Primary Exam Format | Purely subjective essays (800 points across four core blocks) | Objective multiple-choice (MBE, 50%) + subjective essays/tasks (50%) | Short-answer (objective) + essay (subjective) |
| Nature of Passing Standard | Absolute score floor (must reach 400/800 on core subjects) | State-set absolute scaled score (e.g., 266 or 270 out of 400) | Total scaled score cutoff (set annually by the examination commission) |
| Test Equating | No equating mechanism; relies directly on graders’ raw scores | Rigorous equating and scaling; uses anchor items on the MBE to calibrate across administrations | Standardized score conversion (score equating) to eliminate variance across subjects and graders |
| Control of Grading Subjectivity | Double-grading only (third grader if gap is wide); cannot correct for systemic strictness or leniency across an entire grading panel | Essay scores are scaled to the cohort’s MBE distribution (mean and standard deviation) | Statistical conversion formulas standardize raw essay scores across subjects and grading panels |
In educational measurement, setting a fixed passing score rests on one core prerequisite: 400 points in one year must represent the same level of competence as 400 points in another.
The U.S. UBE can enforce fixed passing lines of 266 or 270 because half the exam comes from a nationwide multiple-choice bank (the MBE) equated across years by the National Conference of Bar Examiners (NCBE) using anchor items, and essay scores are statistically scaled to the cohort’s MBE performance. Japan’s bar exam similarly converts raw essay marks into standardized scores to cancel out differences in difficulty and grading severity across subjects and examiners.
(Starting in 2026, U.S. jurisdictions have begun transitioning to the restructured NextGen UBE; the comparison in this section reflects the current UBE.)
In Stage 2 of Taiwan’s bar exam, by contrast, all 800 core points come from open-ended essays written and graded by rotating panels of professors each year. While the Ministry of Examination uses double-grading—calling in a third reader when two graders diverge sharply—that mechanism only catches isolated outliers by an individual grader.
Whenever a year’s prompts skew esoteric, or an entire panel of graders habitually anchors borderline answers at 40 to 45 points, raw scores across the entire cohort sink together. Without score standardization, the full risk of fluctuating question difficulty and grading severity is shifted onto the test-takers.
Across those three years, disqualified candidates formed a self-help association, marched in the streets, and petitioned the Ministry of Examination and the Legislative Yuan, asking: “Does someone with 399 points lack legal competence while someone with 400 points has it?” Lawmakers cited psychometric research at public hearings, challenging the validity and reliability of using raw essay scores as an absolute cutoff when exam difficulty cannot be precisely controlled.
Multiple disqualified candidates also filed collective administrative lawsuits, arguing that the abrupt rule change violated the principles of legitimate expectation and proportionality.
Although the Taipei High Administrative Court (Judgment 108-Su-1236) and the Supreme Administrative Court (Judgment 110-Shang-733) ultimately ruled against the candidates—holding that decisions on examination methods involve professional judgment in selecting licensed professionals, a core area of the Examination Yuan’s examination powers to which courts should extend due deference—years of protests, legislative grilling, and a Control Yuan investigation placed heavy pressure on both examination authorities and legal academia.
The 2021 Reversal: How Did the Threshold Stop Biting Overnight?
Held in late 2021 (ROC Year 110), with results delayed to early 2022 by the COVID-19 pandemic, Stage 2 of the bar exam delivered a sharp reversal. On the day the pass list came out, the number of passers jumped from 650 the previous year to 940—and the Ministry of Examination formally announced that zero candidates in the top 33% had been disqualified by the 400-point threshold.
Not a word of the regulation had changed, yet disqualifications dropped to zero within two years. The Ministry never acknowledged loosening its grading standards, but once 400 points became a hard tripwire, both test-takers and graders adjusted their behavior.
Candidates adapted their test-taking strategies gradually, while a shift in graders’ psychology offers the key hypothesis for why disqualifications suddenly hit zero in 2021. This shift can be viewed from three angles:
1. Defensive Prep by Candidates and Cram Schools (Raising the Core-Subject Floor)
Before 2018, many candidates pursued a “single-subject breakout” strategy: pouring study hours into electives like Intellectual Property or Labor and Social Law to score in the 60s or 70s, using that cushion to offset low marks in weaker core areas like Commercial Law or Public Law.
Once the 400-point floor took effect, a stellar elective score could no longer rescue a sub-400 core total. After two to three years of eliminations, candidates and bar-prep cram schools converged on a highly standardized defensive playbook:
- Abandoning lopsided prep to fortify the four core blocks: Candidates devoted over 90% of their study time to securing baseline competence across Constitutional/Administrative Law, Civil Law/Procedure, Criminal Law/Procedure, and Commercial Law, aiming merely not to fall behind on the elective.
- Standardizing “defensive answer templates”: Cram schools drilled candidates in a floor-protecting essay formula—“issue labeling, multi-view juxtaposition, and personal synthesis.” Candidates start by explicitly tagging the legal issue at stake, lay out competing academic and judicial positions side by side, and close with a balanced concluding view. Even when blindsided by an obscure prompt on exam day, test-takers can follow this scaffolding to fill the expected page count and secure baseline credit from graders, sharply reducing single-subject blowouts.
2. Grade Inflation and Defensive Grading (The Primary Hypothesis for Zero Disqualifications)
Cram schools and candidates, however, had already begun shifting strategies in 2019 and 2020, yet roughly 300 top-third candidates were eliminated in 2020—showing that candidate adaptation alone could not overcome a moving grading yardstick. The available clues point to the decisive variable behind the sudden drop to zero in 2021: a shift in mindset among the professors grading the exams.
Taiwan’s academic legal community is small. Among the hundreds of top-third candidates cut by the 400-point rule between 2018 and 2020 were standout LL.M. students from top law schools—sometimes the very teaching assistants or thesis advisees of the professors sitting on the grading panels. After years of student protests blaming “killer graders” for wiping out entire groups of candidates, alongside Control Yuan investigations and administrative litigation, graders felt the acute weight of deciding candidates’ fates.
By this reasoning, around 2021, question drafters and graders made a crucial self-correction: drafters steered clear of ambush prompts likely to cause mass wipeouts, returning to mainstream doctrinal issues, while grading panels developed a tacit norm of “defensive grading.” So long as an essay showed coherent structure and addressed the key issues, graders aimed to score it at or above the 50% passing line—nudging papers that might previously have received 42 to 45 points up into the 50 to 55 range.
Across an 800-point essay exam, if graders relax their grading by an average of just 5%, the entire cohort’s core-subject totals rise by 40 points. That grade inflation lifted borderline candidates who previously clustered in the 380–395 range straight past the 400-point threshold—offering the most plausible explanation for why disqualifications suddenly dropped to zero in 2021.
3. Why Not Repeal It? A Bureaucratic Equilibrium Where Everyone Saves Face
If the threshold has eliminated virtually no one since 2021, why hasn’t the Ministry of Examination either tightened grading again or repealed the 400-point rule outright? Judged from the standpoint of bureaucratic politics, the status quo is actually an equilibrium that gives every stakeholder an off-ramp:
- For the National Bar Association and senior practitioners: The Ministry can point to the regulations and maintain that “the 400-point threshold remains in force, upholding baseline quality in core legal subjects.”
- For law students and the Legislative Yuan: With annual admissions back above 900 and hundreds of top-third candidates no longer getting disqualified, street protests and mass lawsuits have naturally subsided.
- For examination and judicial authorities: With the Supreme Administrative Court having upheld the constitutionality and legality of the 400-point rule, the administrative discretion and institutional dignity of the agencies remain intact.
Under this tacit understanding, the 400-point threshold has receded from an active screening mechanism into a symbolic line of defense—kept on the books to reassure advocates of supply control, while in practice screening out only extremely low scores.
The Limits of Back-End Gatekeeping—and the Two Paths Ahead
Looking back over half a century of Taiwan’s bar exam—from the martial-law era when only a handful to a few dozen lawyers were admitted each year, to the first major expansion in 1989 after the lifting of martial law, and then to the 33% two-stage reform of 2011—every expansion of the bar has reflected genuine demand for legal talent driven by democratization and commercial growth. Every push to tighten the gates, conversely, has reflected protectionist pushback from incumbent practitioners facing market competition.
This institutional experiment demonstrated a clear lesson: without upstream caps on university law enrollment, a back-end essay cutoff lacking psychometric equating cannot in the long run withstand the adaptive response of both test-takers and the grading ecosystem.
Looking ahead, Taiwan’s legal licensing system faces two broad paths:
The first—and the scenario most likely to continue—is “ritualized coexistence”: the 400-point rule stays on the books, question difficulty and grading remain moderate, and annual admissions hold steadily above 900 new lawyers.
The second path recognizes that if the legal community genuinely wants to fix the supply-demand imbalance and raise practitioner quality, the real battleground is not the 400-point line in Stage 2. Instead, it lies in two long-discussed structural reforms: the Judicial Yuan and Ministry of Examination’s proposal for a “unified legal professional qualification exam (integrating the selection of judges, prosecutors, and attorneys),” and the Ministry of Education’s upstream accreditation and enrollment reductions across university law programs.