Opinion: Oversight Is Not Relocation—Before Abolishing the Control Yuan, Answer: Who Will Supervise the Legislature?

Placing oversight powers into the Legislative Yuan does not automatically create oversight; the essence of separation of powers lies in checks and balances, not in numerical disputes between five-branch and three-branch systems.

On July 29, 2026, the Legislative Yuan's constitutional amendment committee held a public hearing where proposals resurfaced to 'abolish the Examination and Control Yuan, transfer examination authority to the Executive Yuan, and merge oversight powers into the Legislative Yuan.' On July 31, the sixth-term Control Yuan commissioners' term expired. The Taiwan People's Party unveiled a 'Lights-Out Day Memorial Monument' in front of the Control Yuan, advocating 'abolish the institution but not the powers,' returning oversight and audit authorities to the legislature.

Notably, opinions at the hearing were not uniform. Even scholars supporting the abolition of the Control Yuan have warned that if Taiwan adopts a unicameral legislature and directly incorporates oversight powers into the Legislative Yuan, authority could become excessively concentrated.

As of August 1, 2026, there are no sitting Control Yuan members. According to a Central News Agency report on July 31, the agenda includes a vote on the confirmation of new commissioners scheduled for September 29. Petitions from the public can still be received, and staff may conduct preliminary work, but decisions requiring commissioners—such as impeachment, censure, correction, and certain anti-corruption penalties—are suspended. Some cases already before the disciplinary court may face expiration risks due to the lack of a decision-maker on whether to appeal.

This vacancy period places 'abolishing the institution' and 'suspension of oversight' in the same political frame, exposing the real constitutional issue: Taiwan should not debate merely whether the fifth branch of government should be dismantled, but whether any political majority can arbitrarily 'turn off the lights' on oversight.

I do not defend the Control Yuan based on the historical sanctity of the five-power constitution. The political coloration in commissioner nominations, whether case selection is biased, investigation efficiency, and what actual improvements followed corrective actions—all should undergo the strictest scrutiny. However, 'poor performance under the current system' cannot automatically lead to the conclusion that 'transferring power to the Legislative Yuan will make it better.' If the recipient lacks the will to oversee, or is itself the entity being overseen, reform becomes mere bureaucratic relocation.

1. The 'lights out' at the Control Yuan is not reform success, but institutional blackout

The current vacancy does not mean all legal remedies for citizens are interrupted; the Control Yuan is not a court, and citizen petitions are not lawsuits. However, it does mean the constitutional external oversight function cannot make substantive decisions on individual cases—such as impeachment, censure, correction, and certain integrity penalties. Claiming 'petitions are still accepted' equates to 'operations continue normally' is like saying a hospital still functions because registration is open, yet no doctors are available to diagnose or treat.

Taiwan is not experiencing this blackout for the first time. The third-term Control Yuan commissioners' term ended on January 31, 2005, and the fourth-term commissioners were not inaugurated until August 1, 2008—a real vacancy of three and a half years. Judicial Yuan Interpretation No. 632 (2007) stated that the president should nominate timely, and the Legislative Yuan should actively exercise its confirmation power. If any constitutional body passively fails to act, causing the Control Yuan to cease functioning, it undermines the integrity of the constitutional system. The interpretation's rationale further noted that lawmakers could design appropriate legal mechanisms to maintain the Control Yuan's normal operation.

Nearly two decades later, the same flaw reappears. This is the counter-evidence that 'returning all oversight powers to the legislature' must confront: under the current system, if any link—nomination, review, or agenda—fails, oversight powers can halt. If investigative, audit, integrity, and human rights functions are all bundled into the legislative system without stronger term protections, statutory acting mechanisms, budgetary autonomy, and public reporting, does the system enhance checks and balances, or merely hand the switch and spotlight to the same hand?

The essence of oversight is maintaining institutional distance from power, not loyalty to a particular power center. If supervisors must first obtain political permission from those they oversee to decide whom to investigate, for how long, and when to release reports, it is no longer external oversight, but selective self-discipline within power.

2. Three branches are not a math answer; oversight is not a legislative appendage

The value of separation of powers does not lie in the number 'three,' but in functional separation, mutual checks, and the prohibition of self-judgment by power. Democratic nations may adopt a three-branch system or establish central banks, electoral bodies, audit institutions, anti-corruption agencies, and national human rights institutions beyond the three branches. The legitimacy of these bodies should be judged not by counting how many 'yuans' exist, but by whether their functions are necessary, independently exercisable, and democratically accountable.

Advocating abolition because the oversight system originates from ancient imperial censors is a fallacy of origin. Courts, legislatures, and civil service systems also have ancient roots; the legitimacy of modern institutions is not determined by ancestry, but by whether they protect rights, constrain power, and withstand scrutiny today. What should be dismantled are dysfunctional structures, not necessary functions discarded merely because their names are old.

Article 7 of the current Constitution of the Republic of China (Additional Articles) clearly states that the Control Yuan is the highest national oversight body, exercising powers of impeachment, censure, and audit. Judicial Yuan Interpretations No. 325 and No. 585 further explain that the Control Yuan's investigative powers for exercising impeachment, censure, correction, and audit have their own constitutional function. To abolish the Control Yuan or transfer its core powers, the constitutional amendment procedure under Article 12 of the Additional Articles must be followed; it cannot be achieved merely by ordinary laws or organizational acts—'relocation' is insufficient.

The Legislative Yuan's investigative power exists to assist its constitutional duties regarding legislation, budgets, and administrative supervision. The Control Yuan's investigative power serves impeachment, censure, correction, and audit. Interpretation No. 325 stating that investigative power is 'exclusively exercised by the Control Yuan' does not mean the Legislative Yuan has no investigative power. Interpretations No. 585 and Constitutional Court Judgment No. 9 (2024) further confirm that the two bodies' investigative powers differ in authority basis, function, and purpose, and can in principle operate in parallel. However, when investigating the same matter, necessity must be carefully weighed to avoid excessive interference with institutions or individuals. This precisely shows that the two powers can coexist, but cannot be considered fully interchangeable merely by changing signs.

A sharper institutional gap is that Interpretation No. 14, under the current framework, excludes legislators from the scope of oversight by the Control Yuan. This does not mean legislators are entirely unsupervised—criminal justice, elections and recalls, asset declarations, conflict of interest, political donations, and legislative ethics each have their own mechanisms. However, it does mean the Control Yuan cannot use impeachment to pursue legislators' official misconduct. Therefore, if functions originally managed by external bodies—asset declarations, conflict of interest, political donations, and investigations—are reintegrated into the Legislative Yuan, and a body composed of legislators supervises legislators, it becomes easier for the supervised to determine investigation scope, head appointments, budgets, and whether reports are made public.

Legislative self-discipline can supplement external oversight but cannot replace it. Otherwise, we merely transform the concern that 'executive power might influence the Control Yuan' into the reality that 'the legislative majority directly controls oversight power.'

3. Do not reduce 80,000 petitions to political arithmetic

The Taiwan People's Party criticized Control Yuan performance, citing over 80,000 public petitions received by the sixth-term Control Yuan, with only about 1,300 investigations completed. This critique has political impact, but analytically, not every citizen petition should be treated as a mandatory formal investigation.

Statistics released by the Control Yuan in February 2026 show that from August 1, 2020, to December 31, 2025, it received 82,182 citizen petitions, assigned 1,541 for investigation, issued 1,365 investigation reports, issued 481 corrective actions, 1,203 improvement requests, 125 impeachment cases (involving 205 individuals), and 600 penalties under the 'Four Sunshine Laws.' These figures end in December 2025, not the final full-term statistics up to July 31, 2026, and the timeframe must not be omitted when cited.

What should truly be questioned is not simplistic arithmetic like 'petitions divided by number of commissioners,' but: how many cases were closed due to lack of jurisdiction, insufficient evidence, duplicate petitions, or existing alternative remedies? What is the median case processing time? How are major cases selected? Are the same standards applied to officials of different parties and levels? How much concrete improvement followed corrections, and how many problems recurred?

The Control Yuan cannot use 'five-power constitution' to evade performance accountability if it fails to publicly disclose auditable triage criteria, processing timelines, conflict-of-interest rules, minority opinions, and follow-up tracking results. Conversely, critics cannot use incorrect denominators to declare all unassigned petitions as commissioner negligence.

What is truly needed is a public oversight performance dashboard: processing times, correction adoption rates, improvement completion rates, recurrence rates, post-impeachment disciplinary outcomes, and distribution of investigations across power branches should all be regularly published. Oversight must not only report 'how many cases were handled,' but must prove 'what was changed.'

4. International cases prove: the signboard is not the point—independence is

The U.S. Government Accountability Office (GAO) is an independent, nonpartisan agency serving Congress, often called the 'Congressional watchdog.' But it is not a temporary investigative team of the congressional majority.

When the GAO head position is vacant, a bipartisan committee of 10 members—including the Speaker of the House, President pro tempore of the Senate, majority and minority leaders of both chambers, and chairs and ranking minority members of relevant committees—recommends at least three candidates to the President. The President then nominates, and the Senate confirms. The Comptroller General serves a 15-year term and cannot be reappointed. After the previous head's term ended, a pre-designated acting head has led the agency continuously since December 30, 2025, ensuring leadership continuity until a successor is nominated by the President and confirmed by the Senate.

Therefore, citing the GAO cannot involve only extracting the four words 'serves Congress' while omitting the bipartisan nomination, 15-year term, and acting head system—key institutional designs that ensure independence.