The CEC System: A Counterintuitive Design of Dual Executive-Legislative Checks

Taiwan's elections nearly couldn't be held after the CEC went three months with vacant seats.

After the 2024 general election, Taiwan entered a period in which the ruling party held a minority in the legislature. The Legislative Yuan was unable to complete the confirmation process for new Central Election Commission members. Under law, the Premier nominates and the Legislative Yuan exercises its confirmation power — a cycle of nomination → rejection → renomination dragged on for months, and outside observers began to worry: if even the agency responsible for running elections can't be staffed, what happens to the 2026九合一選舉是什麼 (zh: 九合一選舉是什麼)?

Paradoxically, this impasse was not a system failure — it was the system operating exactly as it was designed.

CEC members are nominated by the Premier and confirmed by the Legislative Yuan. This design, from the CEC's initial formation in 1980 through the formal passage of the Central Election Commission Organization Act in 2009¹, has been repeatedly criticized by scholars as "prone to gridlock" and "likely to delay elections" — yet it has never been replaced with a more "efficient" version.

Because it is the last structural safeguard of democratic quality. The entire framework of electoral fairness rests on a counterintuitive design in which dual executive-legislative checks prevent any single party from controlling the process alone — better to stall than to let any one political branch monopolize the institution that runs elections.

Legal Basis and Institutional Framework

The Central Election Commission is an independent agency under the Executive Yuan². An independent agency means that although it sits within the Executive Yuan's organizational structure, it exercises its authority free from the Premier's direction and oversight — placing it in the same category as the National Communications Commission (NCC), the Fair Trade Commission, and the Taiwan Transportation Safety Board.

The Central Election Commission Organization Act was enacted in 2009³, consolidating CEC regulations that had previously been scattered across various executive orders into a single statute. The law stipulates:

• Number of members: 9 to 11, including one chairperson and one vice chairperson
• Term: 4 years, with no possibility of reappointment after the term expires
• Selection process: Nominated by the Premier, appointed after Legislative Yuan confirmation
• Party proportion clause: Members belonging to the same political party may not exceed one-third of the total membership

The "no reappointment" provision is worth noting. Most independent agency members are designed to serve "one renewable term" (e.g., NCC commissioners), but the CEC requires members to leave once their term ends. The reason is to prevent commissioners from currying favor with the ruling party in their final years in office in hopes of reappointment — Taiwan chose to accept the cost of "institutional experience loss" in order to eliminate this structural incentive.

The Member Selection Process: Mutual Checks Between Two Political Branches

The member selection process looks straightforward on paper:

1. The Premier nominates 9–11 candidates
2. The Legislative Yuan exercises confirmation power (requiring a majority vote)
3. Those confirmed are appointed by the President
4. The four-year term begins

In practice, the political tension embedded in this process is far denser than the text suggests.

When the ruling party holds a large majority (controlling both the Executive Yuan and the Legislative Yuan): Nominations pass easily, but the ruling party faces public scrutiny over whether it has nominated too many members of its own party — at this point, the "no more than one-third from the same party" clause becomes the key constraint. During the 2008 Ma Ying-jeou administration and the 2016 Tsai Ing-wen administration, both saw nomination lists challenged for "insufficient cross-party representation"⁴.

When the ruling party holds a minority (the Executive Yuan and the Legislative Yuan controlled by different parties): Nomination gridlock becomes the norm. The Premier nominates people the Legislative Yuan won't confirm; the Legislative Yuan wants people the Premier won't nominate. Cycles of nomination / rejection / renomination can stretch for months. The post-2024 period, in which the Legislative Yuan delayed confirming new CEC members, is a textbook case.

Outside observers are quick to frame CEC gridlock under a minority-ruling-party scenario as "Party A is boycotting, Party B is ramming through" — but from an institutional design perspective, the gridlock itself is part of the design. If the Premier alone could decide CEC membership and the Legislative Yuan could only rubber-stamp, the CEC would become a tool of the ruling party. Conversely, if the legislative majority could decide unilaterally, the CEC would become a tool of the parliamentary majority. Forcing the two branches to compromise on a list is the core of this design.

The cost is "potential gridlock" — and Taiwan has chosen to bear that cost.

From 1980 to 2009: Three Decades of Institutional Evolution

The CEC did not spring fully formed from stone in 1980. Its predecessor can be traced back to the "election committee" structure of the martial law era.

Before 1980: Local elections (county magistrates, county councilors) were administered by local governments; national-level elections (supplementary Legislative Yuan and National Assembly seats) were overseen by the Ministry of the Interior. The entire election administration system was deeply intertwined with the party-state apparatus — the people running elections were people within the ruling party system, and there was no concept of "neutrality" because "neutrality" was defined by the party-state community.

1980: The Central Election Commission was formally established, but it remained under the Ministry of the Interior, and most of its members were appointed through the ruling party system. This version of the CEC functioned more as an "election administration coordination body" than as an "independent election authority" in the modern sense.

The 1990s: A cascade of democratization milestones — the lifting of martial law (1987), constitutional reform (1991–2000), and the first direct presidential election (1996) — forced gradual adjustments to the CEC's structure⁵. Starting in the 1990s, the member selection process began incorporating "cross-party" requirements, though it still lacked a clear statutory foundation.

2009: The Central Election Commission Organization Act was enacted⁶. The CEC was elevated from an "executive-order-level coordination body" to a "statutorily established independent agency" — members now require Legislative Yuan confirmation, a "no more than one-third from the same party" clause was added, and reappointment after a full term was prohibited. This is the CEC as we know it today.

This evolution from party-state co-governance to dual checks took nearly three decades.

What the CEC Does: 99% Technical, 1% Political

The vast majority of the CEC's actual work is technical and legal, with only a tiny fraction involving political judgment. If you haven't heard what the CEC is busy with outside of election night, it's because its work isn't supposed to frequently make the news.

Its primary functions include[^7]:

• Announcing election dates: In accordance with the Civil Servants Election and Recall Act, the CEC issues election announcements a set number of days before each election
• Vetting candidate qualifications: Reviewing whether registered candidates meet legal requirements (age, residency, education, whether their political rights have been stripped, etc.)
• Producing election bulletins (選舉公報): Compiling candidate information and policy platforms, printing and mailing them to every household
• Polling station allocation: Coordinating with local government civil affairs departments on polling station locations and ballot distribution
• Vote tabulation and winner announcements: Compiling nationwide vote counts on election night and announcing the winners
• Referendum proposal review: Following amendments to the Referendum Act (2017, 2019, 2022), responsible for reviewing petition thresholds and setting referendum dates
• Recall case review: Procedural review of recall proposals, petition stages, and voting dates

Behind each of these tasks lies a thick layer of regulations and technical specifications, but none involve value judgments about "who would be better to win." To give a concrete example: the font size, line spacing, candidate photo dimensions, and character limits for policy platforms in an election bulletin are all specified in writing — the CEC's job is to enforce these rules so every candidate gets the same amount of space, not to judge whose platform is "better." Similarly, the distance between polling stations, the cap on ballots per station (typically 3,000), and the vote-counting process (callers, recorders, and monitors staffing each station) all have SOPs — the CEC's job is to make those SOPs work correctly across 17,000+ polling stations simultaneously, which is itself a massive coordination challenge.

The CEC's Division of Labor With Other Agencies

The CEC does not run elections alone. The entire election system is a cross-agency collaborative network[^8]:

• Ministry of the Interior, Department of Household Registration: Voter rolls (who can vote and where)
• Ministry of the Interior, National Police Agency: Polling station security (election-day police deployment, election dispute prevention)
• Ministry of the Interior, Investigation Bureau: Election fraud (vote-buying investigations, evidence for election litigation)
• Control Yuan: Political donations (Political Donations Act reporting and auditing)
• Local government civil affairs departments: Polling station operations (station chiefs, administrators, and monitors at each location)

Within this network, the CEC plays the role of "rule-maker and referee": setting rules, vetting qualifications, and announcing results — but not directly administering elections. The people actually working at polling stations are civil servants and public school staff seconded by local civil affairs departments.

Structural Controversies: Recurring Every Four Years

The CEC's institutional design has several structural controversies that have recurred since its establishment:

1. Is the party proportion clause (same party ≤ 1/3) effective?

Although commissioners cannot exceed one-third from the same party, the political leanings of "independent" (no-party-affiliation) commissioners are difficult to quantify. Critics argue the clause is only cosmetic balance — the Premier can nominate candidates who are "formally independent but substantively aligned." Supporters counter that at least the law draws a line, which is better than no line at all.

2. Nomination gridlock under minority-ruling-party conditions

The 2008 Ma Ying-jeou administration leaned toward pan-Blue dominance, the 2016 Tsai Ing-wen administration leaned toward pan-Green dominance, and the post-2024 period has seen nomination gridlock under a minority ruling party — this pattern is not an accident of any single term but evidence that the system itself is sensitive to political structure.

When gridlock occurs and vacancies exceed a certain proportion, the CEC can continue operating with its remaining members (the Organization Act has a fallback minimum membership provision), but the legitimacy of its decisions will be questioned. In an extreme scenario, if gridlock persists until the date when an election announcement should be issued, it could theoretically affect the election calendar. Taiwan has not yet reached this extreme, but this tail risk has always existed.

3. Where is the boundary of the CEC's "proactivity"?

When election disputes arise (e.g., disinformation, AI deepfakes, foreign interference), should the CEC "proactively intervene" or "passively refer to the relevant agencies"? Current practice leans toward the latter — when the CEC identifies a problem, it refers it to prosecutors, the NCC, the Ministry of Justice, etc., rather than acting as a "truth arbiter" itself. This conservative approach reduces the risk of the CEC being politicized, but it also means response times to new forms of election interference may be slower than the public expects⁷.

4. No reappointment after a full term: the cost of institutional experience

"No reappointment after a full term" ensures that structural incentives are severed, but it comes at the cost of institutional experience loss. A CEC commissioner spends four years learning the ins and outs of referendum petition threshold calculations, cross-party candidate dispute resolution, and election-night vote tabulation workflows — then must leave at term's end — and the next cohort of incoming commissioners has to learn everything from scratch. This trade-off has long been debated in academia: would "one renewable term, but reappointment requires a cross-party majority vote" be a better alternative? No consensus has yet formed.

International Comparison: Taiwan's CEC in the Asian Context

Viewing Taiwan's CEC in an international context reveals that its design is relatively sophisticated by Asian standards.

US FEC (Federal Election Commission): 6 commissioners, legally allocated 3 to each of the two major parties, nominated by the President and confirmed by the Senate. The problem is that the "3 each" design means major decisions require 4 votes → the commission frequently deadlocks 3-3, and has been criticized as "designed to be paralyzed"⁸. Over the past decade-plus, the FEC has on multiple occasions been unable to rule on major election funding cases due to vacancies and unresolved gridlock — this is the consequence of a "symmetrical checks" design taken to its extreme. Taiwan's CEC chose the "executive nomination + legislative confirmation" asymmetric path precisely to avoid this kind of symmetrical paralysis.

Japan: No central election commission. National elections are coordinated by the "Election Division" under the Ministry of Internal Affairs and Communications, while actual election administration is handled by prefectural election management committees (都道府県選挙管理委員会)⁹. Japan's election committee members are elected by prefectural assemblies, making the system more decentralized than Taiwan's. The advantage is decentralization and reduced risk of a single agency being politicized; the disadvantage is difficulty in cross-prefecture coordination and low uniformity in national election standards, requiring the Ministry of Internal Affairs and Communications to do extensive horizontal coordination work during every national election.

South Korea: The National Election Commission (중앙선거관리위원회) has 9 commissioners, with the President, the Chief Justice of the Supreme Court, and the National Assembly each nominating 3¹⁰. This "one-third from each branch of government" design is more sophisticated in its checks and balances — executive, legislative, and judicial sources ensure that no single political branch can dominate alone. In practice, however, nominations from the judicial branch are also frequently questioned as to whether they are "truly politically neutral," since the Chief Justice is nominated by the President and confirmed by the legislature, making the judicial source indirectly subject to political influence.

India: The Election Commission of India is extremely independent, with autonomous budgeting and personnel authority to the point that "even the President cannot interfere with its day-to-day operations"¹¹. The Indian Election Commission gradually expanded its powers after the 1990s, even gaining the authority to order the suspension of cabinet ministers and the reassignment of civil servants during election periods, making it one of the most powerful election bodies in the world. The advantage is blocking political interference; the disadvantage is that excessive independence brings controversy over "the commission itself becoming too powerful and lacking accountability" — can you check a body that checks everyone?

Taiwan's CEC — "executive nomination + legislative confirmation + party proportion clause + no reappointment" — represents a middle path between the US FEC's tendency toward gridlock and India's excessive independence. It is not perfect, but within Asian democratic systems, it is a design that has been repeatedly weighed — maintaining relative independence while leaving final personnel authority to be jointly determined by two democratically legitimate political branches.

The 2026 Election: The CEC's Work Timeline

The 2026九合一選舉是什麼 (zh: 九合一選舉是什麼) voting day is November 28, 2026. The CEC's work timeline is roughly as follows[^14]:

• 2026/08/20: Election announcement issued
• 2026/08/29–09/04: Candidate registration period (with minor variations by office type)
• Before 2026/10/16: Candidate list finalized
• 2026/10/23: Candidate ballot number lottery
• Early 2026/11: Election bulletin printing and mailing
• 2026/11/28: Election day
• 2026/11/28 evening: Vote tabulation and winner announcements
• From 2026/12: Election litigation accepted (the CEC does not adjudicate but handles administrative procedures)

This timeline looks almost boringly routine — and that is precisely where the CEC's value lies. When an election schedule can be laid out months in advance, precise to the day; when every candidate knows what will happen next; when every voter receives the same bulletin and walks into a polling station built to the same specifications — the legitimacy of the election is upheld by what lies behind that boredom.

The CEC in the Age of AI Deepfakes and Disinformation

During the 2024 general election, AI-generated candidate deepfake videos became an election issue for the first time¹². The CEC's current response strategy remains primarily "referral to relevant agencies": deepfake content is reported to the NCC and prosecutors, foreign interference is reported to the Ministry of Justice Investigation Bureau, and the CEC itself does not conduct fact-checking or content removal.

This conservative approach has its institutional logic — once the CEC steps in to make judgments about "what is disinformation," it will be drawn into political attacks accusing it of "suppressing one party's speech on behalf of another." But the cost is slower response times and shifting the burden of judgment to other agencies.

Finding a new balance between "maintaining neutrality" and "responding in real time to new forms of election interference" is the CEC's greatest institutional challenge in the coming years.

Conclusion: Better to Stall Than to Be Monopolized

If you remember only one thing: the CEC's institutional design was never about "running elections as efficiently as possible" — it was about making it as difficult as possible for any single political force to monopolize the running of elections.

Premier nominates, Legislative Yuan confirms, no more than one-third from the same party, no reappointment after a full term — these four provisions check each other, forcing any political force that wants to manipulate the CEC to pass through both the executive and legislative lines of defense simultaneously. This design has been controversial since the CEC's initial formation in 1980, yet it has never been changed.

Because it is the last structural safeguard of democratic quality. In the post-2024 scenario of a minority ruling party and nomination gridlock, the cost of this safeguard is clearly visible: yes, there will be gridlock; yes, there will be political tug-of-war; yes, it is not efficient. But compared to letting the institution that runs elections be controlled by any single party, that cost is worth paying.

Electoral fairness is not an empty phrase — it is upheld by these counterintuitive, mutually checking, would-rather-stall institutional designs.

Politics Hub · 2026 九合一選舉 (zh: 2026 九合一選舉) · 政治獻金透明度 (zh: 政治獻金透明度) · 九合一選舉是什麼 (zh: 九合一選舉是什麼) · 選舉公報 (zh: 選舉公報) · Democratization