TW-O-01: Taiwan Megatrends — The 2030s Questions

Status: DRAFTWords: 9,222

1. Key Takeaways

The cross-Strait question is not one question among several; it is the master variable through which every other trajectory in this document runs. The semiconductor industry's location decisions, the energy grid's blockade mathematics, the conscription pool's demographic arithmetic, the budget battles of the divided legislature, and the diplomatic-ally count are all, at bottom, sub-questions of whether the People's Republic of China attempts to compel unification in the 2027–2035 window and whether deterrence holds (TW-F-02, TW-F-04, CN-E-02). This document therefore inverts the usual megatrends format: Section 2 states the master variable first, and every subsequent section is read partly as an input to it. The honest baseline is irreducible uncertainty — the corpus's discipline is to refuse both the complacent reading (deterrence is stable because conflict would be irrational) and the alarmist reading (a dated invasion timeline), and to track indicators instead.
The "2027 window" is a capability milestone, not a known deadline — and conflating the two readings is the most common analytical error the corpus documents. Admiral Philip Davidson's March 2021 testimony that the PLA threat could be "manifest" within six years, and reported US intelligence assessments that Xi Jinping has instructed the PLA to be ready by 2027 (the PLA's centenary), describe a capability-construction schedule, not a decision to act [TBD-VERIFY: the precise Davidson wording and the Burns/CIA characterisations of the 2027 instruction]. CN-I-02 §5 supplies the strongest counter-evidence to deadline readings: the 2023–2025 purge cascade through the Rocket Force, the defence-minister succession, and the Central Military Commission itself suggests a party leadership that does not yet trust the instrument it has built. The window framing survives because it is the planning-prudent reading; the corpus treats the PLA as an untested force commanded by a leadership with unverified confidence in it.
The likelier coercion pathway is grey-zone strangulation, not D-Day — and Taiwan's defences are least developed exactly there. The Joint Sword (2024) and Strait Thunder (2025) exercise series rehearsed blockade and quarantine geometries rather than amphibious assault; the subsea-cable cuts, the ADIZ normalisation, the law-enforcement "inspection" framings around Kinmen, and the economic-coercion instruments (CN-E-02, TW-F-04 §10) constitute a pressure architecture designed to compel negotiation below the threshold that triggers US intervention. A quarantine enforced by coast guard rather than navy poses the defining 2030s dilemma: Taiwan cannot shoot first at a customs inspection, and the United States may not treat one as the casus foederis an invasion would be. The energy-storage mathematics of Section 6 — LNG measured in days, not months — is the quarantine scenario's critical multiplier.
The silicon shield is real but eroding at both ends, and the 2030s will test whether it was ever load-bearing. Taiwan's roughly 90 per cent share of advanced-node foundry output (TW-G-02, TW-G-03) deters in two directions — making destruction costly to Beijing and protection valuable to Washington — but the Arizona/Kumamoto/Dresden dispersal architecture, accelerated by Trump-2 tariff pressure and the March 2025 USD 100 billion expansion pledge, moves leading-edge capacity offshore on a 2028–2032 schedule, while PRC indigenisation (SMIC's 7nm, the Huawei Ascend line) erodes the monopoly from the other side. The corpus's three-account discipline (TW-G-03 §12) applies: dispersal as insurance that preserves Taiwanese centrality, dispersal as slow-motion shield transfer, and the AI-demand surge as a force that may make the question moot by making all capacity scarce. Which account proves true is among the highest-stakes industrial-policy questions on earth.
Taiwan's democracy enters the 2030s having survived its hardest decade since democratisation — but the stress test is structural, not episodic. The 2024–2026 sequence — a three-way presidential race producing a president with 40 per cent and a hung Legislative Yuan, the May 2024 legislative-powers expansion and the Bluebird counter-mobilisation, the Constitutional Court's October 2024 ruling and the subsequent justice-confirmation paralysis that left the court below quorum, the failed mass-recall campaigns of 2025, and the 2026 budget confrontation (TW-I-01 §6, TW-D-04, TW-D-06) — demonstrated both resilience (no actor defected from constitutional forms) and dysfunction (core institutions can be disabled by ordinary-law manoeuvre). The 2030s question is whether divided government becomes a workable norm, a paralysing one, or a vector through which united-front pressure operates inside the system rather than against it.
The demographic cliff is Taiwan's least discussed structural threat and the one most certain to arrive on schedule. Taiwan's total fertility rate — [TBD-VERIFY: approximately 0.87 in 2023–2024, contesting with South Korea for the world's lowest] — has the population shrinking since 2020 and becoming super-aged (over-65 share above 20 per cent) in 2025. The defence implication is arithmetic: the conscription cohort that the January 2024 twelve-month service restoration depends on (TW-F-04 §8) shrinks every year, the reserve base shrinks behind it, and the care economy competes with the armed forces and the fabs for the same dwindling young workforce. Migration policy — historically restrictive, quietly liberalising — becomes a national-security variable. Depopulation does not change whether Taiwan is defensible in 2027; it changes whether the defence model is sustainable in 2035.
The energy system is the single most exploitable vulnerability in any coercion scenario, and the 2025 nuclear endpoint sharpened rather than settled the question. Taiwan imports roughly 97 per cent of its energy; LNG storage covers [TBD-VERIFY: on the order of 8–14 days at normal consumption], coal several weeks; the nuclear-free-homeland policy completed with Maanshan Unit 2's May 2025 shutdown even as the August 2025 restart referendum showed a majority for reversal that failed the turnout threshold (TW-G-05). The renewables build-out has run persistently behind target while AI-era electricity demand accelerates ahead of forecast. Section 6's scenarios are deliberately coupled to Section 2's: in the deterred-status-quo world, energy is an economics problem; in the quarantine world, it is the clock by which Taipei's negotiating position decays.
Taiwan's international space has a hard floor and a soft ceiling, and both are moving. The formal-recognition floor — twelve diplomatic allies after Nauru's January 2024 switch [TBD-VERIFY: current count as of 2026] — continues to erode toward a plausible single-digit endpoint, while the unofficial relationship lattice (the US Taiwan Relations Act architecture, the Japan and EU thickening, the Lithuania precedent and its costs) has grown denser than at any point since 1979 (TW-F-01, TW-N-01). The PRC pressure instruments are evolving from ally-poaching toward systemic lawfare — the campaign to read UN General Assembly Resolution 2758 as having settled Taiwan's status — and toward conditioning third countries' wartime behaviour in advance. The "porcupine plus friends" strategy's limit is that friends are a wasting asset if the friends' own deterrence calculations change.
Four equilibria frame the 2030s: deterred prosperity, pressured resilience, coerced accommodation, and conflict. The first extrapolates the 1996–2026 record — pressure without rupture, prosperity with anxiety. The second is the corpus's modal reading of the near term: intensifying grey-zone coercion absorbed at rising cost. The third arrives not by invasion but by accumulation — energy exposure, demographic exhaustion, allied distraction, and internal polarisation converging until a Taiwanese electorate chooses accommodation as the least-bad option. The fourth is the catastrophe scenario, on which this document refuses false precision: the corpus assigns it no probability and instead specifies the discriminating indicators (Section 8). The base rate counsels neither comfort nor fatalism: the Strait has been "about to erupt" for seventy-five years and has not, and the conditions of the 2030s resemble no previous decade of those seventy-five.
The honest meta-finding: Taiwan's trajectories are less in Taiwan's hands than any comparable polity's, and yet Taiwanese agency keeps proving decisive at the margin. The 1996–2026 record shows external structure (US-China relations, PLA modernisation, semiconductor economics) setting the field, and Taiwanese choices — democratisation itself, the Sunflower rejection of dependency (TW-B-05), the asymmetric-defence turn, the chip-investment discipline — repeatedly determining where on the field events land. Every section of this document therefore ends with indicators split between what Beijing and Washington do and what Taipei does, because the 2030s will be decided in all three capitals at once.

2. The Cross-Strait Question as Master Variable

2.1 The Trend

The structural trend of the 1996–2026 period is the inversion of the military balance and the exhaustion of the political formulas that once managed its consequences. In 1996, when the PLA bracketed Taiwan's first direct presidential election with missile tests, two US carrier battle groups settled the matter; the corpus's military-balance record (TW-F-04) documents the thirty-year programme Beijing built in response — a navy larger by hull count than the US Navy, the world's densest ground-based missile force, and an exercise architecture that since August 2022 (TW-C-06) treats the median line as void and Taiwan's surrounding waters as a rehearsal space. Simultaneously, the political track decayed: the 1992 Consensus formula that underwrote the Ma-era détente (TW-E-01, TW-K-03) was foreclosed on the Taiwan side by the Sunflower Movement and the 2016–2024 elections, and on the Beijing side by Xi Jinping's January 2019 conflation of the Consensus with the one-country-two-systems formula (TW-K-01) — a formula Hong Kong's 2019–2020 trajectory then discredited with Taiwanese voters beyond recovery. What remains in 2026 is the configuration TW-F-02 calls the deterrence-dependence equilibrium: no political process, a military balance shifting against Taiwan, and stability resting on deterrence alone — US capability and ambiguity, Taiwanese asymmetric defence, and Beijing's own cost calculus.

The "2027–2035 window" discourse that dominates planning in all three capitals requires careful decomposition, because it bundles two distinct claims. The capability-milestone reading: Xi has instructed the PLA to be ready to take Taiwan by 2027 — the PLA's centennial — with 2035 the milestone for full modernisation; Admiral Philip Davidson's March 2021 Senate testimony ("the threat is manifest... in the next six years") gave the window its popular name [TBD-VERIFY: precise Davidson wording; the CIA's public characterisation of the 2027 readiness instruction, attributed to Director Burns in February 2023]. The deadline reading: that 2027 (or 2035, or Xi's actuarial horizon) is a decision point at which action becomes probable. The corpus treats the first as well-evidenced and the second as unsupported — readiness instructions are what militaries are for, and no public evidence shows a decision to use force on any date. CN-I-02 §5 supplies the strongest standing counter-evidence to deadline determinism: the purge cascade of 2023–2026 — two consecutive defence ministers expelled, the Rocket Force command replaced wholesale, Central Military Commission members suspended and removed — describes a leadership conducting loyalty surgery on the very force a deadline reading assumes is being honed for use. An army being purged is an army not trusted; an army not trusted is, on the historical record of party-armies, an army the party hesitates to bet its existence on. The inference runs both ways, however: purges can also be read as pre-conflict discipline-tightening, and the corpus declines to resolve what Beijing itself may not have resolved.

2.2 The Corpus Evidence

Three evidence streams anchor the section. First, the grey-zone normalisation record (CN-E-02; TW-F-04 §4, §10): the Joint Sword-2024A/B and Strait Thunder-2025A/B exercises rehearsed encirclement, blockade-checkpoint, and energy-interdiction geometries; PLA air incursions across the former median line became daily; subsea cables to Matsu and around the main island suffered cuts attributed to PRC-linked vessels [TBD-VERIFY: attribution status of the 2023–2025 cable incidents]; and the China Coast Guard asserted "inspection" jurisdiction in Kinmen waters after the February 2024 speedboat deaths. The pattern is a ladder being built rung by rung — each exercise converts a wartime act into a peacetime precedent. Second, the deterrence-investment record (TW-F-04 §5, §8; TW-D-06): Taiwan's defence budget rose from NT$327 billion (2017) toward a proposed NT$949 billion, with Lai Ching-te committing to exceed 3 per cent of GDP in 2026 and signalling a path toward higher shares under US pressure [TBD-VERIFY: the specific Lai 2025–2026 commitments on the 3.32 per cent figure and any 5-per-cent-by-2030 pledge]; conscription returned to twelve months in January 2024; and the asymmetric procurement pipeline (HIMARS, Harpoon coastal defence, the Hai Kun submarine programme) advanced against the counter-pull of prestige-platform preferences. Third, the US-commitment record (TW-F-01): four Biden statements that the US would defend Taiwan, each walked back to strategic ambiguity by staff; the Trump-2 administration's transactional framing — chip-investment and burden-sharing demands running alongside arms sales — introducing a commitment volatility that Beijing probes and Taipei hedges against. The US variable is the deterrence equation's least stable term precisely because it resets every four years.

2.3 The Scenarios

Deterred status quo. Deterrence holds through the window: the PLA's readiness remains unverified and Beijing's risk tolerance bounded; the US relationship, however transactional, never signals abandonment at a moment of PRC confidence; Taiwan's asymmetric investments raise the visible price of invasion beyond what an untested force can promise its commander-in-chief. Pressure continues — exercises, lawfare, economic coercion — but as substitutes for action rather than preparations for it. This is the 1996–2026 record extrapolated, and it is the scenario every other section's optimistic branch assumes.

Coercive-absorption pressure. The likelier dark scenario is not invasion but strangulation: a quarantine or inspection regime enforced by coast guard and maritime militia, energy-import interdiction calibrated below the military threshold, cable cuts and cyber operations against the grid, secondary pressure on shipping insurers and trading partners, and a political demand — negotiations on Beijing's terms — presented as the exit. The scenario's power lies in its alliance-splitting design: a customs inspection is not Pearl Harbor, and the US response to slow strangulation is genuinely uncertain. Taiwan's counter is endurance (energy storage, stockpiles, civil resilience — Sections 5 and 6) and internationalisation (making quarantine costly to third parties whose ships and chips transit the Strait). The blockade-versus-quarantine distinction matters: a declared blockade is an act of war that clarifies allied obligations; a "quarantine" administered as law enforcement is engineered to keep them ambiguous.

Conflict. The catastrophe scenario — amphibious invasion, decapitation strike, or escalation from a grey-zone incident neither side intended — is the one on which this document refuses false precision. The corpus assigns no probability. What it records instead: the war-gamed outcomes are catastrophic for all parties under most assumptions [TBD-VERIFY: CSIS January 2023 wargame headline findings]; the escalation-by-accident pathway grows mechanically with the tempo of grey-zone contact; and the historical base rate of leaders launching wars their militaries privately doubted is not zero. The conflict scenario's most underweighted variant is the limited one — seizure of Kinmen, Matsu, or Pratas as a fait accompli testing whether Washington and Taipei will escalate over outlying islands the US has never committed to defend.

2.4 Indicators to Watch

(1) PLA purge trajectory (CN-I-02): continued senior turbulence signals untrusted instrument; a stabilised, exercise-intensive command cohort signals the opposite. (2) Exercise evolution: the shift from rehearsal to no-notice posture, sustained encirclement beyond one week, or live interdiction of a named vessel — each a rung. (3) The US commitment signal under each administration: arms-delivery backlog clearance versus drift; any explicit conditioning of defence on economic concessions. (4) Taiwan's defence spending share and asymmetry ratio (missiles and mines versus prestige platforms) in each budget cycle (TW-D-06's successors). (5) PRC mobilisation infrastructure: civilian ro-ro ferry militarisation, blood-supply and hospital-capacity surges in the eastern theatre, strategic-reserve accumulation [TBD-VERIFY: the open-source indicator literature on each]. (6) The Kinmen-Matsu legal regime: any PRC declaration converting "inspection" practice into asserted routine jurisdiction. (7) Beijing's domestic framing: whether unification rhetoric moves from patient-inevitability to time-bound necessity in authoritative party texts. None of these indicators is individually decisive; the discipline is to watch for correlated movement across them — and to distrust anyone offering a date.

3. The Silicon-Shield Question

3.1 The Trend

The silicon-shield doctrine — that Taiwan's centrality to advanced semiconductor manufacturing deters attack by making destruction intolerably costly to everyone, Beijing included, and Taiwan's survival strategically valuable to Washington — hardened from metaphor into operating assumption during the 2020–2022 chip shortage, when TSMC's roughly 90 per cent share of advanced-node (sub-7nm) foundry output became the most quoted statistic in geoeconomics (TW-G-01, TW-G-02, TW-N-01 §4). The 2030s question is whether the shield survives its own success. Three forces work on it simultaneously. Geographic dispersal: the Arizona architecture (Fab 21 Phase 1 N4 production from November 2024; Phase 2 targeting N3 around 2028; Phase 3 targeting N2 around 2030, folded into the March 2025 pledge that raised TSMC's announced US commitment to roughly USD 165 billion), Kumamoto's JASM fabs (Fab 1 opened February 2024; Fab 2 targeted 2027), and Dresden's ESMC (ground broken September 2024) move meaningful capacity — though not the leading edge's centre of gravity — offshore on a 2028–2032 completion schedule (TW-G-03 §3–5). PRC indigenisation: SMIC's 7nm production without EUV, the Huawei Ascend accelerator line, and the state-funded self-sufficiency drive erode the monopoly from below, with the US export-control regime (the October 2022 / October 2023 / December 2024 BIS cycles, TW-G-03 §6) acting either as containment or as accelerant — the corpus carries both readings. Demand transformation: the AI build-out has converted advanced logic and packaging (CoWoS) into the binding constraint of the world's largest capital-expenditure cycle, deepening Taiwan's indispensability in the near term even as it supplies the political motive for dispersal.

3.2 The Corpus Evidence

TW-G-03's three-account discipline (§12) is this section's spine, and the evidence cuts across all three accounts. For shield-maintained: the leading edge stays home — N2 mass production began in Taiwan (Hsinchu and Kaohsiung) [TBD-VERIFY: the N2 volume-production timeline, reported for late 2025, and the A16 successor node's 2026–2027 schedule], TSMC's R&D core remains in Hsinchu by policy and law (the cross-Strait and outbound technology-control regimes, TW-G-03 §10), and the overseas fabs run one to two generations behind by design. The "N-minus-one" convention — overseas fabs lag the domestic frontier — has been Taiwanese industrial policy's quiet red line, though Trump-2 pressure tested it [TBD-VERIFY: whether the Arizona Phase 3 N2 timeline narrows the lag below one generation at steady state]. For shield-diluted: the trajectory matters more than the snapshot; if Arizona reaches N2 by 2030 and subsequent phases track the frontier, the US no longer needs Taiwan for its most critical supply by the early 2030s — exactly the window's far end — and a Washington that has onshored its insurance may discount Taiwan's strategic value at the margin that deterrence decisions are made on. The Trump-2 tariff campaign (TW-D-03) made the bargaining explicit: chip investment as the price of security partnership. For shield-obsoleted: the PRC path — if indigenisation reaches good-enough parity for military and AI purposes, Beijing's cost calculus loses its self-deterrence term regardless of where TSMC builds; alternatively, a technology discontinuity (advanced packaging, chiplets, or a post-silicon architecture) could devalue the monopoly the shield rests on. The talent pipeline is the slow variable beneath all three: Taiwan's engineering-graduate cohort shrinks with Section 5's demography, the fabs already recruit from Southeast Asia and India, and the diaspora fabs compete for the same Taiwanese process engineers — [TBD-VERIFY: scale of TSMC's outbound engineer rotations to Arizona/Kumamoto and domestic attrition figures].

3.3 The Scenarios

Shield maintained. The N-minus-one discipline holds; Taiwan retains the frontier (A16 and beyond), the R&D core, and the packaging bottleneck; overseas fabs absorb political pressure while deepening, not diluting, partner dependence on Taiwanese process knowledge — every Arizona wafer still runs on Hsinchu-trained engineering. On this account dispersal strengthens deterrence: allies acquire a stake in TSMC's survival as a company, which is inseparable from Taiwan's survival as a polity. The scenario requires sustained Taiwanese policy discipline (technology-control law, R&D tax architecture, talent investment) and a US administration content with insurance rather than relocation.

Shield diluted. The frontier gap narrows offshore generation by generation; the 2030s end with advanced capacity tri-located (Taiwan, US, Japan), Taiwan's share of leading-edge output falling from ~90 per cent toward [TBD-VERIFY: industry projections clustering around 50–60 per cent by the early-to-mid 2030s]; and the deterrence externality decays quietly — no announcement, just a Washington whose Taiwan contingency planning notes, year by year, that the chip argument weighs less. Taiwanese leverage shifts from monopoly to mere excellence, which is a normal country's asset, not a shield. This is the base-rate scenario if current build schedules complete.

Shield obsoleted. Either PRC good-enough self-sufficiency or an architecture transition removes the monopoly's strategic salience within the decade. The shield was always a by-product, never a strategy Taiwan could file a patent on; this scenario is the reminder that deterrence borrowed from economics is deterrence on a depreciating asset. Taiwan's hedge is to convert chip centrality into institutionalised security relationships while the leverage lasts — the chip-diplomacy doctrine TW-G-03 §11 documents.

3.4 Indicators to Watch

(1) The node gap: the generation distance between Taiwan's frontier and the most advanced overseas TSMC fab, checked at each capacity announcement — the single cleanest shield metric. (2) Where A16 and the post-A16 node site their first volume fabs. (3) CoWoS/advanced-packaging geographic distribution, the bottleneck the node metric misses. (4) PRC indigenisation milestones: SMIC yield at 5nm-class without EUV, domestic EUV-equivalent lithography progress [TBD-VERIFY: status], Huawei Ascend volume against export-control leakage. (5) Taiwanese technology-control legislation and its enforcement against talent and IP outflow. (6) The US posture: whether CHIPS-successor policy and tariff bargaining treat Taiwan as partner or as supplier to be onshored away from. (7) TSMC's domestic capital-expenditure share — the company's own revealed belief about where its future sits.

4. The Democratic-Resilience Question

4.1 The Trend

Taiwan's democracy is simultaneously the most consolidated in Asia by most indices and the most stressed by design: it is the only democracy whose principal adversary claims sovereignty over it, operates inside its information space in its own language, and maintains relationships with one of its two major parties. The 2024–2026 period subjected this system to its hardest compound test since democratisation, and TW-I-01 §6 documents the constitutional mechanics in full. The sequence: the January 2024 election produced a president (Lai, 40.1 per cent in a three-way race) without a legislative majority for the first time since 2004–2008 (TW-D-01); the KMT-TPP bloc passed the May 2024 legislative-powers expansion, triggering the Bluebird movement — the largest street mobilisation since the Sunflower occupation (TW-B-05) — and a Constitutional Court ruling in October 2024 striking the core provisions; the bloc responded by raising the Constitutional Court's quorum and blocking justice confirmations, leaving the court unable to rule [TBD-VERIFY: the court's exact bench count through 2025–2026 and the quorum statute's thresholds]; the 2025 mass-recall campaign against KMT legislators failed across the board (TW-D-04), recalibrating Lai toward cohabitation management (TW-D-05); and the FY2026 budget battle (TW-D-06) carried the confrontation into the defence-spending question itself. The structural finding is double-edged. Resilience: every actor stayed inside constitutional forms — courts were obeyed, recalls were conducted lawfully and their failures accepted, no party called crowds against institutions. Dysfunction: the same period proved that core institutions (the Constitutional Court above all) can be disabled by ordinary legislative majorities, and that a constitution amendable only by near-impossible thresholds (TW-I-01's central theme) cannot self-repair.

4.2 The Corpus Evidence

Three further evidence streams. The recall/referendum volatility: Taiwan's direct-democracy instruments — lowered to accessible thresholds in 2016–2017 — have become continuations of electoral war by other means: the 2018 and 2021 referendum waves, the Han Kuo-yu recall (2020), the 2025 mass-recall campaign, and the August 2025 nuclear referendum (TW-G-05 §6) together establish a system in permanent campaign, with each cycle consuming governing bandwidth and sharpening polarisation. The disinformation defence (TW-N-01 §3): Taiwan is the world's most-targeted information environment by most measures [TBD-VERIFY: the V-Dem/DoubleThink Lab foreign-disinformation rankings], and its civil-society response — the g0v ecosystem, Cofacts, rapid-response government communication, media-literacy curricula — is studied globally as the democratic counter-model to censorship; yet the 2024–2026 cycles showed the threat migrating from fabricated content toward amplified authentic division (recall campaigns, energy fears, "US scepticism" narratives) where fact-checking has no purchase. The identity consolidation: the long NCCU Election Study Center series shows exclusive Taiwanese identification rising from under 20 per cent (1992) to a stable majority — [TBD-VERIFY: the current figures, with Taiwanese-only identification in the low-to-mid 60s per cent and dual identification around 30 per cent in recent waves] — with unification support in low single digits across all age cohorts. This is the deepest structural fact in Taiwanese politics: it forecloses voluntary absorption on any horizon Beijing has articulated, and it frames the KMT's generational question — a party whose founding identity commitments poll weakest among voters under forty, which must either complete its evolution into a Taiwan-centric conservative party or watch the TPP and successor formations harvest its base. The KMT's post-2024 trajectory — competitive in legislative and local politics (TW-D-04, TW-D-05), structurally weak presidentially — keeps both readings alive.

4.3 The Scenarios

Resilient adaptation. Divided government becomes a norm the system metabolises: cross-party convention or statutory repair restores the Constitutional Court; the recall and referendum instruments are re-calibrated (thresholds, frequency limits) by negotiated reform; the 2028 cycle produces either a majority government or a cohabitation with working rules; and the defence consensus survives partisan alternation — the test being whether a future KMT administration sustains the 3-per-cent-plus trajectory it contested in opposition. Taiwan's democratic brand, already its principal soft-power asset (TW-N-01 §3), compounds.

Polarised paralysis. The 2024–2026 pattern extrapolates: each presidential-legislative mismatch reproduces institutional siege, budget hostage-taking, and court vacancy; direct democracy runs as permanent campaign; defence and energy policy lurch with each cycle; and public trust drains symmetrically. Paralysis is not collapse — Taiwan governed adequately through 2024–2026 — but in the master variable's shadow it is strategically expensive: a Taipei that cannot pass a defence budget on schedule is a datum in Beijing's patience calculus and Washington's reliability assessment.

United-front erosion. The adversarial variant: PRC influence operations move from outside the system to inside it — local-faction financing, temple and business networks, retired-officer channels, platform algorithms (the TikTok question), and the cultivation of "war-scepticism" framings that convert deterrence spending into a partisan position [TBD-VERIFY: prosecution counts under the Anti-Infiltration Act 2020–2026]. The scenario's end state is not a pro-unification majority — the identity data forecloses that — but a Taiwan too internally divided to invest in its own defence or to respond coherently under quarantine pressure: erosion as preparation of the battlefield. The discipline here is to avoid the mirror-image error of treating all opposition to DPP policy as united-front product; the corpus's three-account practice exists precisely to keep legitimate KMT/TPP positions and PRC operations analytically distinct.

4.4 Indicators to Watch

(1) The Constitutional Court: restoration to working strength, and how — negotiated repair signals adaptation; continued vacancy signals paralysis. (2) The 2028 presidential-legislative configuration, and whether any cohabitation develops formal working rules (a cohabitation accord, regularised party-leader summits). (3) Defence-budget votes under divided government — the cleanest fusion indicator with Section 2. (4) The NCCU identity and unification-preference series, annually [TBD-VERIFY: series continuity]. (5) KMT leadership succession and platform evolution: the cross-Strait plank at each party congress, and the age profile of its vote. (6) Anti-Infiltration Act and political-finance prosecutions — both their incidence and their evidentiary quality. (7) Referendum/recall frequency per cycle, as a volatility metric. (8) Civic-trust and willingness-to-defend surveys [TBD-VERIFY: the INDSR/Academia Sinica willingness-to-fight series], the deepest resilience measures available.

5. The Demographic-Cliff Question

5.1 The Trend

Taiwan's fertility collapse is the structural trend with the least uncertainty attached: the children who will not be conscripted, employed, or taxed in 2040 have already not been born. The total fertility rate has run below replacement since 1984 and below 1.0 since the early 2020s — [TBD-VERIFY: ~0.87 in 2023–2024 on Ministry of the Interior data, contesting with South Korea (~0.72–0.75) for the world's lowest; the "world's lowest fertility" label applies to Taiwan in some years and series but not all]. The population peaked at roughly 23.6 million around 2019–2020 and has declined since; deaths have exceeded births every year since 2020; the National Development Council's projections put the 2040 population in the low 22 millions and falling, with the over-65 share crossing 20 per cent ("super-aged") in 2025 and approaching 30 per cent by the late 2030s [TBD-VERIFY: NDC medium-variant figures]. Pro-natalist policy across four administrations — childcare subsidies, parental leave expansion, fertility-treatment coverage — has moved the curve not at all, which matches the East Asian policy record generally; the binding constraints (housing costs, working hours, gendered care burdens, and the marriage rate that mediates almost all Taiwanese fertility) lie deeper than subsidy reaches.

5.2 The Corpus Evidence and the Defence Arithmetic

The corpus's defence record makes the demographic question a security question. The January 2024 conscription extension to twelve months (TW-F-04 §8) restored training depth to a force whose conscript cohort shrinks mechanically: the male cohort reaching conscription age falls from roughly [TBD-VERIFY: ~110,000 in the mid-2020s toward ~75,000–80,000 by the mid-2030s on cohort arithmetic from registered births]. The volunteer force persistently misses recruitment targets [TBD-VERIFY: the 2023–2025 staffing-rate figures, reported in the low 80s per cent of authorised strength]; the reserve system being rebuilt around the 2022-established All-Out Defense Mobilization Agency draws on the same shrinking pool; and the civil-defence ambition (TW-F-04 §9) requires exactly the young volunteers the care economy and the fabs also bid for. Depopulation thus enters the master variable twice: it thins the defence base on a schedule Beijing can read as well as anyone, and it may inform Beijing's own patience — a Taiwan that ages into dependency is, on one united-front reading, a Taiwan that eventually negotiates [TBD-VERIFY: PRC commentary explicitly making this argument]. Against this, the same arithmetic runs harder in China, whose fertility and ageing trajectory is comparably severe at vastly greater scale — demography disadvantages both sides of the Strait, and the net strategic effect is genuinely unclear. The economic face: the working-age population has fallen since 2015; the migrant workforce (~800,000, predominantly Southeast Asian, concentrated in manufacturing, fishing, and care [TBD-VERIFY: current stock]) fills the floor of the labour market while immigration policy for mid-skill settlement remains restrictive; the long-term-care system (the Long-Term Care 2.0 framework) and the pension funds face the standard super-aged squeeze, with the labour-insurance fund's actuarial exhaustion repeatedly projected for the early 2030s absent reform [TBD-VERIFY: latest actuarial date]. Housing cost — among the world's highest price-to-income ratios in Taipei [TBD-VERIFY: ratio] — operates as both a fertility suppressant and a standing political grievance for the under-forty cohort whose loyalty every party now courts.

5.3 The Scenarios

Managed shrinkage. Taiwan does what no East Asian polity has yet done at scale: treats migration as strategy — settlement pathways for mid-skill workers and care labour, aggressive talent recruitment (including from Hong Kong and the Taiwanese diaspora), automation across manufacturing and care, and a defence model (drones, autonomous systems, a professionalised smaller force with a genuinely trained reserve) that decouples deterrence from cohort size. Fertility stays low; the country adapts around it. The technology trajectory of Section 3 is this scenario's enabler — Taiwan is better positioned than any peer to automate its way through depopulation.

Drift. The base rate: pro-natalist spending continues to fail politely; migration liberalises incrementally and insufficiently; the conscription pool, the volunteer shortfall, and the care gap all worsen on schedule; pension reform is deferred past each election; and by the mid-2030s the demographic constraint binds visibly on the defence budget (personnel costs crowding procurement) and on growth. Nothing breaks; everything tightens.

Compound squeeze. The interaction scenario: demographic decline meets a Section 2 contingency or a Section 3 shield-dilution — an ageing, shrinking Taiwan with a thinning defence base, rising care costs, and reduced chip leverage faces intensified coercion with weakened endurance. This is less a separate scenario than the mechanism by which demography feeds the coerced-accommodation equilibrium of Section 8.

5.4 Indicators to Watch

(1) Annual births — the single hardest number in this document; any sustained recovery above [TBD-VERIFY: the ~135,000 recent floor] would be regionally unprecedented. (2) Migration policy: any settlement-pathway legislation for mid-skill workers, and the migrant-care workforce's growth rate. (3) Military staffing: volunteer-force fill rates, conscript cohort size, and reserve-training throughput, annually. (4) Personnel share of the defence budget — the crowding-out metric. (5) Pension and labour-insurance reform: enacted, or deferred again. (6) The under-forty housing-and-wages grievance in each electoral cycle, the channel through which demography becomes politics.

6. The Energy-Security Question

6.1 The Trend

Taiwan runs one of the world's most advanced industrial economies on an islanded grid fed almost entirely by sea: roughly 97 per cent of energy supply is imported, there are no pipelines and no interconnectors, and the storage buffers are measured in days and weeks (TW-G-05 §2). The policy arc of the past decade compounded the exposure. The 2016 "nuclear-free homeland by 2025" commitment — Fukushima-era politics codified (TW-G-05 §4) — was implemented essentially on schedule: the fourth nuclear plant was mothballed and its restart rejected by the 2021 referendum, and Maanshan Unit 2, the last operating reactor, shut down in May 2025, taking nuclear's share from ~12 per cent (2016) to zero. The replacement strategy — gas as bridge, renewables as destination — delivered the gas (LNG's share rose past 40 per cent) but not the renewables: the 20-per-cent-by-2025 target was missed by a wide margin [TBD-VERIFY: the realised renewables share, reported around 10–12 per cent in 2024–2025], with offshore wind advancing impressively by global standards but behind its own auction schedule, and ground-mounted solar colliding with land scarcity and agricultural politics. The reversal arc then began (TW-G-05 §6): the 2024–2026 KMT-TPP legislature amended the law to permit reactor-life extension; the 23 August 2025 referendum on restarting Maanshan returned a large Yes majority that failed the turnout threshold [TBD-VERIFY: the vote and turnout figures], leaving restart legally possible, politically signalled, and practically unresolved — Taipower's restart feasibility studies, fuel procurement, and regulatory re-licensing put any returned reactor in the late 2020s at the earliest.

6.2 The Corpus Evidence and the Blockade Mathematics

The strategic core is the storage arithmetic. LNG — now the marginal fuel for both power and industry — can be stored on the order of [TBD-VERIFY: 8–14 days at normal consumption, with statutory targets to raise this toward 20+ days via the Guantang/Taichung/Pingtung terminal build-out]; coal stocks run several weeks [TBD-VERIFY: ~30–42 days]; oil reserves are the deepest at 90-plus days under the stockpiling law. In any Section 2 quarantine scenario, LNG carriers — enormous, slow, uninsurable under threat, and owned by third parties — stop sailing first, possibly on insurance-market signal alone before a single shot. The grid's decay curve under interdiction is therefore the quarantine scenario's clock: rationing within days, industrial shutdown within weeks, and the political pressure of a darkened, de-aired-conditioned electorate operating on Beijing's side of the negotiation. The demand side tightens simultaneously: the AI build-out's electricity appetite — TSMC alone consumes [TBD-VERIFY: ~8–9 per cent and rising of national electricity] — collides with the supply transition, and Taipower's financial position (accumulated losses from absorbing imported-fuel costs [TBD-VERIFY: scale]) constrains the investment the build-out requires. The 2026 budget battle carried an energy front for exactly this reason (TW-D-06 §9). The defence-policy convergence is notable: energy resilience — storage, dispersed generation, microgrids, hardened substations — now appears in defence-resilience planning (TW-F-04 §9) as squarely as munitions stockpiles, and grid-node sabotage and cyber operations feature in every serious contingency analysis.

6.3 The Scenarios

These are deliberately coupled to Section 2's. In the deterred-status-quo world — energy as economics: the question is cost and carbon; the build-out proceeds (offshore wind, storage, terminals), nuclear returns partially as a pragmatic baseload-and-resilience hedge with cross-party cover from the referendum's Yes majority, and the AI demand is met expensively but adequately. In the pressured world — energy as endurance: storage targets, demand-rationing plans, and restart decisions are made explicitly on blockade mathematics; nuclear's case strengthens decisively (a fuelled reactor is eighteen months of stored energy in a building), and the indicator of seriousness is whether storage and resilience spending is justified in security language in budget documents. In the coercion world — energy as the clock: an interdiction-shaped crisis arrives before the buffers are built, and the grid becomes the mechanism by which a quarantine converts into a negotiation Taipei enters weak. The scenario discipline: Taiwan's energy choices cannot change Beijing's behaviour, but they directly set the price and duration of the coercion option — energy policy is deterrence policy conducted by other ministries.

6.4 Indicators to Watch

(1) LNG storage days-of-cover, the single most strategic number in Taiwanese infrastructure, against the statutory targets and terminal completion dates. (2) Any Maanshan (or Kuosheng) restart decision, and whether it is framed as energy economics or national security. (3) The renewables trajectory against revised targets, especially offshore-wind completions and grid-scale storage. (4) Taipower's balance sheet and tariff politics — the financial feasibility constraint. (5) AI-datacentre and fab demand growth against capacity additions; any TSMC statement on power as a siting constraint. (6) Resilience spending: microgrids, substation hardening, fuel-stockpile legislation. (7) Any PRC exercise explicitly rehearsing energy interdiction — the indicator that fuses this section back into Section 2.

7. The International-Space Question

7.1 The Trend

Taiwan's external existence runs on two tracks whose divergence is the trend: the formal-recognition floor keeps falling while the unofficial-relations ceiling keeps rising. The floor: diplomatic allies fell from 22 (2016) to 12 after Nauru's January 2024 switch [TBD-VERIFY: the count as of 2026, and any subsequent losses], the residue concentrated in the Pacific, the Caribbean and Central America, Eswatini, and the Holy See — the last the only ally with systemic weight, and itself the object of continuous PRC negotiation. The trajectory points toward single digits in the 2030s, and Taipei's revealed strategy has shifted from chequebook retention toward managed decline: each loss is absorbed, condemned, and de-dramatised. The ceiling: the unofficial lattice has thickened beyond anything since derecognition (TW-N-01 §2, TW-F-01) — the US relationship's post-2016 institutional deepening; Japan's "Taiwan contingency is a Japan contingency" turn and its 2025 sharpening [TBD-VERIFY: the Takaichi-era formulations and the PRC response]; the European movement from indifference to chip-and-democracy engagement (the Lithuania representative-office episode of 2021 and the PRC economic coercion it drew); parliamentary diplomacy at scale (IPAC); and the New Southbound web of economic relationships (TW-E-02). The PRC instrument set is evolving in response: from ally-poaching (a depleting resource) toward systemic lawfare — the campaign to establish that UN General Assembly Resolution 2758 (1971) settled not merely China's seat but Taiwan's status, contested by US and allied legislatures from 2023–2025 [TBD-VERIFY: the resolutions and statements rejecting the conflation] — alongside organisation-by-organisation exclusion enforcement (no WHA invitation since 2017, ICAO, Interpol, the renaming pressure on airlines and corporations) and the pre-conditioning of third countries' crisis behaviour through economic dependence.

7.2 The Corpus Evidence

TW-N-01's stacked-frames analysis is the evidence spine: Taiwan's international standing now rests on the chip frame (Section 3's leverage), the democracy-beacon frame (Section 4's brand), and the flashpoint frame (Section 2's salience) — and all three are wasting or double-edged assets. The chip frame depreciates with the shield (Section 3); the beacon frame depends on the democratic performance Section 4 interrogates; the flashpoint frame buys attention at the price of being discussed as a problem between great powers rather than a polity with agency — the framing TW-N-01 §7 identifies as what external perception misses. The participation record shows the ceiling's height and its limits: observer or member status in APEC, the WTO ("Separate Customs Territory of Taiwan, Penghu, Kinmen and Matsu"), and the Asian Development Bank under negotiated names; exclusion everywhere statehood is the ticket; the CPTPP application (September 2021, days before the PRC's own) parked in a consensus-rule limbo that makes Taiwan's accession hostage to every member's China relationship [TBD-VERIFY: the accession working-group status as of 2026]. The "porcupine plus friends" strategy — asymmetric self-defence plus a lattice of partners with stakes in Taiwan's survival — is the synthesis Taipei has converged on across administrations, and its limit is structural: friends without obligations are friends whose help is recalculated in every crisis, and the lattice's load-bearing strand remains a single US commitment of deliberate ambiguity (TW-F-01). The Ukraine demonstration runs both ways in Taiwanese debate: proof that democracies will arm and sustain a victim of aggression, and proof that they will not fight for one.

7.3 The Scenarios

Thickening compensation. The floor falls to single digits and stops mattering: the unofficial lattice continues to densify — trade agreements (the US-Taiwan 21st-century initiative's successors, a UK/EU investment thread, CPTPP accession in a post-China-application equilibrium [TBD-VERIFY]), defence-industrial co-production, intelligence and coast-guard cooperation — until Taiwan's functional position resembles a non-recognised ally. The scenario's driver is the master variable: partners invest in Taiwan to the degree they fear the alternative. Squeeze. The PRC succeeds in converting economic leverage into behavioural pre-commitment: allies-of-allies decline transit visits, the 2758 lawfare hardens into UN-system practice, the remaining diplomatic allies fall in a clustered run timed for psychological effect, and the unofficial lattice's growth stalls as middle powers de-risk their China relationships. Taiwan is not absorbed but is progressively unmentioned — the international-space dimension of coerced accommodation. Status crisis. A Section 2 contingency forces every state's position from constructive ambiguity into declared policy within days — recognition, abandonment, or armed support — and the international-space question is settled not by trend but by event, in whichever direction the event runs.

7.4 Indicators to Watch

(1) The ally count, with attention to clustering — single losses are noise; a timed run is signal. (2) The Holy See relationship [TBD-VERIFY: the Sino-Vatican bishops-agreement renewal cycle], the one switch with systemic meaning. (3) The 2758 lawfare ledger: each parliament or government formally rejecting (or adopting) the PRC reading. (4) CPTPP accession movement, either direction. (5) WHA observer status — the annual, legible barometer of PRC flexibility. (6) The Japan and Philippines policy trajectories, the two neighbours whose contingency posture most directly conditions US options. (7) Whether new unofficial agreements include continuity-of-relations or crisis-consultation language — the quiet measure of whether "friends" are becoming something more.

8. Synthesis — Four Equilibria for 2030s Taiwan

The six questions above are inputs to a single synthesis, and the synthesis must be handled with the gravity the subject demands: unlike the other megatrends documents in this corpus, one of Taiwan's candidate equilibria involves the deliberate destruction of the polity under study. The corpus frames four equilibria. They are scenarios, not predictions; no probabilities are assigned, because none can be honestly derived; the indicator sets are what distinguish them in real time.

Equilibrium 1 — Deterred prosperity (the extrapolation scenario). Deterrence holds visibly enough that the pressure architecture plateaus; the silicon shield is diluted but the economy compounds on AI demand; divided government matures into workable cohabitation; the energy build-out and a partial nuclear return close the blockade-mathematics gap; demography tightens everything without breaking anything; the unofficial lattice thickens. Taiwan in 2040 is richer, older, smaller, still anxious, still free. Signature indicators: PLA purge turbulence continuing; exercises plateauing in scale and pattern; defence budgets passing on schedule under both parties; LNG cover above 20 days; the node gap holding at N-minus-one; the NCCU identity series stable. This is the 1996–2026 record extrapolated, and the base rate behind extrapolation is the strongest argument it has.

Equilibrium 2 — Pressured resilience (the modal near-term reading). Coercion intensifies along the grey-zone ladder — quarantine rehearsals becoming quasi-permanent presence, cable and cyber attrition, economic and lawfare squeeze — and Taiwan absorbs it at rising cost: defence and resilience spending crowding other priorities, insurance and shipping premia taxing the economy, each electoral cycle fought partly over how much endurance to fund. The system holds because Sections 3–7's investments hold. The distinction from Equilibrium 1 is tempo and cost; the distinction from Equilibrium 3 is that absorption capacity grows as fast as pressure. Signature indicators: exercise tempo rising but never crossing into sustained interdiction; resilience budgets justified in security language and passed; willingness-to-defend polling stable or rising under pressure; the lattice's crisis-consultation language thickening.

Equilibrium 3 — Coerced accommodation (the accumulation scenario). No invasion occurs; none is needed. The vectors converge — energy exposure unclosed, demographic exhaustion thinning the defence base, shield dilution discounting Taiwan in Washington, polarised paralysis disabling response, allied distraction (a Middle East or European crisis, a transactional US administration, a Japan that blinks) — until a Taiwanese government, or electorate, concludes that negotiation on Beijing's schedule is the least-bad path: a peace framework, an interim arrangement, a "diplomatic truce" that hardens into trajectory. The corpus notes the precedent vocabulary already exists (the 2008–2016 détente's architecture, TW-E-01, TW-K-03) and that the identity data (Section 4) makes consent to absorption implausible — which is why this equilibrium's mechanism is not conversion but resignation, and why its early indicators are about morale and hedging, not opinion on unification. Signature indicators: willingness-to-defend polling falling; capital flight and emigration of the young rising [TBD-VERIFY: baseline emigration and second-passport data]; defence budgets failing or shrinking in real terms; energy-storage targets missed without political consequence; a mainstream presidential candidacy running explicitly on a negotiated framework and winning.

Equilibrium 4 — Conflict (the catastrophe scenario). Blockade-to-war escalation, a limited offshore-island seizure that does not stay limited, an invasion attempt in the window's later years, or an accident on a ladder rung neither capital intended. The corpus refuses false precision here deliberately: war-game literature, intelligence commentary, and the deterrence theory canon all converge on wide uncertainty bands, and a governance corpus adds no value by pretending otherwise. What it can say: the scenario's likelihood is endogenous to everything above — every indicator in Sections 2–7 is also an input to this one — and its consequences would be of an order that reorganises every other document in this corpus and most of the global economy with it [TBD-VERIFY: Bloomberg Economics and CSIS estimates of global-GDP impact of a Strait conflict, commonly cited around 10 per cent of global GDP]. Signature indicators: the Section 2.4 mobilisation set moving in correlation; sustained interdiction of named vessels; PRC domestic framing shifting to time-bound necessity; US commitment signalling collapsing at a moment of PRC confidence — and the honest acknowledgment that surprise is itself a documented feature of war initiation.

What distinguishes them early. Three crosscutting indicator pairs carry the most discriminating power before 2030. First, the purge-exercise pair (CN-I-02, CN-E-02): command turbulence plus plateauing exercises points to Equilibria 1–2; command stabilisation plus exercise escalation is the most dangerous combination on the board. Second, the budget-endurance pair: whether Taiwan's defence and resilience budgets pass, grow, and shift asymmetric under divided government (TW-D-06's successors), and whether the LNG-storage and grid-hardening targets are met — jointly the cleanest measure of whether Taipei is buying endurance faster than Beijing is buying coercion. Third, the morale pair: the identity series and the willingness-to-defend series — stable identity with falling willingness is the specific signature of the accommodation path, and it is measurable years before any negotiation begins. An observer who tracks only these three pairs through 2027–2030 will know most of what this document can responsibly teach.

9. Conclusion

The six questions of this document are one question asked six ways: can a polity of twenty-three million — ageing fastest, governing itself under divided institutions, running the world's most strategically coveted industry on an islanded grid, recognised by almost no one — sustain the conditions of its own continuity through a decade in which its adversary's declared patience and demonstrated capability are converging? The 1996–2026 record gives a double answer, and the corpus's discipline is to hold both halves. Taiwan has been far more durable than its structural exposure suggests: it democratised under missile bracketing, built the most consequential industrial monopoly of the digital age while formally unrecognised, absorbed the 2022 and 2024–2025 exercise escalations without panic or capitulation, and ran a constitutional stress test through 2024–2026 in which every actor, however bitterly, stayed inside the rules. And Taiwan has been more fragile than its performance suggests: its deterrence is substantially borrowed, its energy endurance is measured in days, its defence demography is a closing window, and its political system has shown it can disable its own constitutional court over a partisan quarrel while a carrier group exercises off its east coast.

The 2030s will be decided by which of those two Taiwans compounds faster — and, more than for any other country in this corpus, by decisions taken in capitals that are not its own. That asymmetry is the reason this document has been stricter than its siblings about the limits of foresight: scenarios, not predictions; indicators, not prophecy; and on the conflict question, no probabilities at all, because none can be honestly derived and false precision in this domain is not merely an analytical error but a moral one. The four equilibria of Section 8 are falsifiable frames, and the indicator pairs attached to them are the falsification apparatus. This document should be revisited after each presidential cycle (2028, 2032, 2036), after each major PRC party congress and exercise-pattern shift, and immediately upon any event that moves a tail scenario toward the centre — a sustained interdiction, a clustered recognition run, a restart or abandonment decision on the nuclear question, a node-gap collapse, or a willingness-to-defend inflection. Until then, the extrapolation equilibrium remains the way to bet, the pressured-resilience reading the way to plan — and the questions remain open, which is, for Taiwan above all, the point: an open question is the status quo, and the status quo is what twenty-three million people have repeatedly voted to keep.

Primary Sources Consulted:

Bush, Richard C., Difficult Choices: Taiwan's Quest for Security and the Good Life (Brookings, 2021); Uncharted Strait: The Future of China-Taiwan Relations (Brookings, 2013).
Rigger, Shelley, Why Taiwan Matters: Small Island, Global Powerhouse (Rowman & Littlefield, 2011) and The Tiger Leading the Dragon (2021).
Davidson, Admiral Philip, testimony before the US Senate Armed Services Committee, 9 March 2021 [TBD-VERIFY: exact transcript wording of the six-year assessment]; subsequent INDOPACOM and DNI Annual Threat Assessment statements on PLA 2027 readiness, 2021–2026.
US Department of Defense, Military and Security Developments Involving the People's Republic of China (annual reports to Congress, 2020–2025 editions).
CSIS (Center for Strategic and International Studies), The First Battle of the Next War: Wargaming a Chinese Invasion of Taiwan (January 2023); CSIS ChinaPower project reports on blockade and quarantine scenarios (2024) [TBD-VERIFY: titles and headline findings].
Taiwan Ministry of National Defense, National Defense Report and Quadrennial Defense Review editions 2017–2025; the FY2025–FY2026 defence-budget submissions and Legislative Yuan review records.
Lee Hsi-min and Eric Lee, writings on the Overall Defense Concept and asymmetric defence, 2019–2025; INDSR (Institute for National Defense and Security Research) annual assessments.
National Chengchi University Election Study Center, Taiwanese / Chinese Identity and Unification–Independence Preference trend series, 1992–2026 [TBD-VERIFY: latest wave figures].
TSMC investor communications: quarterly earnings calls, the November 2024 Arizona Fab 21 production announcement, the 3 March 2025 US-expansion announcement, and annual reports 2020–2025; Taiwan MOEA statements on outbound-technology controls.
Miller, Chris, Chip War: The Fight for the World's Most Critical Technology (Scribner, 2022), and subsequent commentary on the dispersal architecture.
SEMI, TrendForce, and Boston Consulting Group / SIA reports on advanced-node capacity distribution and projections, 2021–2026 [TBD-VERIFY: specific editions and share projections].
Taiwan National Development Council, Population Projections for the Republic of China (Taiwan) (2022 and 2024 editions); Ministry of the Interior vital-statistics series.
Taiwan Ministry of Economic Affairs and Taipower: Energy Statistics Handbook, LNG-storage statutory targets and terminal schedules, and the August 2025 referendum record (Central Election Commission).
Central Election Commission, official results: 2024 presidential and legislative elections; the 2025 recall votes; the 23 August 2025 referendum.
Taiwan Constitutional Court, Judgment 113-Hsien-Pan-9 (October 2024) on the legislative-powers expansion [TBD-VERIFY: judgment number], and the subsequent confirmation-and-quorum record.
DoubleThink Lab, V-Dem Institute, and IORG research on PRC information operations against Taiwan, 2019–2026.
Glaser, Bonnie S., Jessica Chen Weiss, and Thomas Christensen, "Taiwan and the True Sources of Deterrence," Foreign Affairs (2023/2024), and the surrounding deterrence-reassurance debate.
Blanchette, Jude, and Ryan Hass, The Taiwan Long Game and related Brookings/CSIS work on the 2027 window discourse, 2023–2026.
Mainland Affairs Council and Straits Exchange Foundation records; PRC Taiwan Affairs Office statements and the 2022 White Paper The Taiwan Question and China's Reunification in the New Era.
Reuters, Financial Times, and Nikkei Asia reporting on the Joint Sword-2024A/B and Strait Thunder-2025 exercises, the subsea-cable incidents, the Nauru switch (January 2024), and the 2025–2026 budget battles.
Kyodo and Japanese government records on the Japan-Taiwan contingency formulations, 2021–2026 [TBD-VERIFY: the 2025 statements and PRC response].
Bloomberg Economics, "The Price of a Taiwan Conflict" analyses [TBD-VERIFY: edition and figures]; Rhodium Group, The Global Economic Disruptions from a Taiwan Conflict (2022).

Related Documents:

TW-D-06: Taiwan's 2026 Budget Battle, the KMT-TPP Bloc Confrontation, and the Defence-Spending-to-Three-Per-Cent-of-GDP Question (2025–2026)
TW-F-02: Cross-Strait Dynamics 2000–2026 — From Chen Shui-bian's Independence Rhetoric to Lai Ching-te's Deterrence-Dependence Equilibrium
TW-F-04: The Cross-Strait Military Balance, Asymmetric Defence, and the Overall Defense Concept (2017–2026)
TW-G-02: Taiwan's Semiconductor Ecosystem Beyond TSMC — UMC, MediaTek, Realtek, ASE, and the Industrial Policy Architecture (1987–2025)
TW-G-03: Taiwan's Semiconductor Diaspora, TSMC's Global Fab Architecture, and Chip Diplomacy (2020–2026)
TW-G-05: Taiwan Energy Policy, Nuclear Politics, and the Security-of-Supply Question (1978–2026)
TW-I-01: The ROC Constitutional Architecture — The Five-Yuan System, the Seven Amendment Rounds, and the Politics of a Constitution That Can No Longer Be Amended (1947–2026)
TW-J-01: Three Accounts — The 28 February 1947 Incident and Its Memory Politics (1947–2026)
TW-K-03: The 7 November 2015 Ma Ying-jeou – Xi Jinping Singapore Meeting Decision (2014–2026)
TW-N-01: Taiwan in International Perceptions — Chip Island, Beacon, and Flashpoint (1971–2026)
TW-A-01: back-reference added by symmetry sweep
TW-H-PRES-04: back-reference added by symmetry sweep
TW-F-03: The New Southbound Policy and Taiwan's Southeast Asia Turn
TW-M-01: Taiwanese Identity and the Status-Quo Doctrine — The Island's Organising Ideas
TW-J-02: the 19 May 1949 Martial Law Declaration and the World's Longest Martial Law at Its 1987 Lifting, the Punishment of Rebellion Act and the Communist-Spy Statute's Informer-Reward Machinery, the Military Tribunals That Tried Civilians for Four Decades, the 1950s Leftist Eradication and the Green Island Prison World, the Lei Chen and Peng Ming-min Cases, the 1979 Kaohsiung Incident and the Making of the Formosa Generation, the Terminal Murders of the 1980s, and the Contest Among the State-Violence/Transitional-Justice Account, the Cold-War-Necessity Account, and the Complicated-Society/Grey-Zone Account Cross-Country Related Documents:
CN-E-02: PRC Cross-Strait Policy in 2025 — Strait Thunder-2025A/B, Lai's "Foreign Hostile Force" Designation, and Trump-2 Bargaining
CN-I-02: The People's Liberation Army and Party-Military Relations — "The Party Commands the Gun" from the Jiang–Hu Handover to the Xi-Era Purges (2002–2026)