The Hollowing Out of Asia's Factory Floor: Japan, South Korea, and Taiwan Confront Industrial Decline

For decades, Japan, South Korea, and Taiwan formed the manufacturing backbone of the global economy — producing the steel, ships, cars, chips, and consumer electronics that powered postwar growth. That era is fraying. Factory activity across all three economies contracted for much of 2025, with Japan's new orders declining for two and a half consecutive years and Taiwan's manufacturing PMI spending ten straight months below the 50-point threshold that separates expansion from contraction [1]. South Korea's industrial output growth slowed to just 0.5% in 2025, its weakest since the COVID-19 contraction of 2020 [2].

The question facing policymakers in Tokyo, Seoul, and Taipei is whether this represents a manageable transition — shedding low-margin commodity manufacturing to climb the value chain — or a deeper structural rot that threatens the economic foundations of all three nations.

The Numbers: How Deep Is the Decline?

Japan's manufacturing sector accounted for about 20.6% of GDP in 2023, according to World Bank data — essentially flat from 20.8% in 2010 [3]. The stability of that headline figure, however, masks real deterioration in output volumes and employment. Japan's GDP contracted 1.8% in the third quarter of 2025, with exports of goods and services shrinking 1.2% quarter-on-quarter as U.S. tariffs hit the export-oriented economy [4]. Japan's working-age population has fallen 16% from its 1995 peak of 87.3 million to 73.7 million in 2024, a demographic headwind with no parallel in the Western deindustrialization of the 1980s [5].

Source: World Bank Open Data

Data as of Dec 31, 2023CSV

South Korea's manufacturing value added as a share of GDP fell from 29.0% in 2011 to 25.5% in 2023, before recovering slightly to 26.6% in 2024 — a decline of roughly 2.4 percentage points over the period [3]. More telling is the erosion in global market share. A June 2025 analysis found that all eight of South Korea's major industries — semiconductors, displays, smartphones, automobiles, petrochemicals, steel, shipbuilding, and batteries — lost global market share simultaneously for the first time in a decade [6].

Source: World Bank Open Data

Data as of Dec 31, 2024CSV

The sector-specific numbers are stark. Korean display makers Samsung and LG held 98.5% of the small-to-medium OLED market in 2015; by early 2025, that had fallen below 60% as Chinese competitors BOE and CSOT expanded [6]. The secondary battery industry's global share collapsed from 34.7% in late 2020 to 18.7% by early 2025, with all three major Korean battery makers — LG Energy Solution, Samsung SDI, and SK On — posting operating losses [6]. Steel exports fell 8.8% in 2025, petrochemical exports dropped 11.4%, and shipbuilding's share slid from 30% to 17% [6][2].

Taiwan presents a bifurcated picture. The island's economy grew nearly 9% in 2025, driven almost entirely by AI-fueled semiconductor exports [7]. But traditional manufacturers — metals, plastics, machine tools — were left behind. One machine toolmaker reported overall exports down 30% [8]. Taiwan's manufacturing PMI spent January through November 2025 in contraction territory before barely crossing into expansion at 50.9 in December [9].

Comparing to Western Deindustrialization

The pattern bears structural similarities to the U.S. and U.K. experience in the 1980s and 1990s, when manufacturing's share of GDP fell sharply — from roughly 20% to 12% in the U.S. between 1980 and 2000, and from 25% to 15% in the U.K. over a similar period [10]. But the Asian trajectory differs in key respects. Japan's manufacturing share has remained relatively stable around 20% for over a decade — far higher than today's U.S. figure of roughly 11%. South Korea's share, while declining, still exceeds 25%. The contraction here is less about absolute deindustrialization and more about margin compression and competitive displacement, particularly by China.

Japan's auto sector, which employs 5.58 million people — 8.3% of the total workforce — illustrates the vulnerability. Trade and tariff pressures have squeezed earnings without triggering the mass layoffs seen in the U.S. Rust Belt, in part because Japan's tight labor market (2.5% unemployment in 2025) means firms hoard workers rather than shed them [4][5].

Source: World Bank Open Data

Data as of Dec 31, 2025CSV

The China Factor: Overcapacity Meets Structural Weakness

China's state-subsidized industrial overcapacity is the single largest external pressure on these economies. China's steel sector suffers from chronic overproduction, its electric vehicle industry has an estimated 5–10 million units of annual excess capacity, and its export machine has hollowed out competitors across Southeast Asia — 4,300 Thai factories closed in just two years, across furniture, electronics, garments, auto parts, and steel [11].

The European Parliament published a study in 2026 documenting how China's decentralized governance structure — where local officials are evaluated on GDP growth — created systematic duplicate investment across provinces, resulting in overcapacity that is then dumped on global markets [12]. For South Korea, the effects are direct: chip exports to China fell 14.6% year-on-year in May 2025, while petrochemical exports dropped 11.4% and general machinery exports declined 13.6% [6].

But attributing all the pain to Chinese competition would be incomplete. Japan's 350 business bankruptcies in 2024 attributed to labor shortages point to a domestic structural crisis that has nothing to do with Beijing [5]. South Korea's "triple crisis" — weakening exports, flagging domestic demand (constant-value retail sales fell 0.2% year-on-year through early 2025, continuing a three-year downturn), and talent constraints — reflects internal imbalances as much as external pressure [6]. Taiwan's traditional manufacturers face labor shortages and high costs that preceded any Chinese overcapacity surge [13].

The honest assessment is that China is accelerating a decline that domestic factors had already set in motion. Aging populations, rising wages, and insufficient R&D reinvestment in legacy sectors created the conditions; Chinese competition exploited them.

The Policy Response: Billions for Chips, Less for Everyone Else

Governments have responded with massive industrial policy interventions — but almost all the money flows to semiconductors.

Japan has allocated approximately $65 billion in semiconductor subsidies between 2021 and 2025, representing 0.71% of GDP — a higher proportion than the U.S. CHIPS Act (0.21% of GDP) [14]. TSMC's Kumamoto facility received $8 billion in subsidies, with a second fab pledged $4.9 billion more [15]. The government committed $12 billion to Rapidus, the domestic startup aiming to mass-produce 2nm chips by 2027 in Hokkaido [16]. Between May 2020 and March 2022, Japan's METI also subsidized 439 onshoring projects covering medical equipment, auto parts, electronics, and semiconductors [17].

South Korea announced a 33 trillion won ($23.25 billion) semiconductor support package in April 2025, including 20 trillion won in low-interest loans to chip companies through 2027 [18]. The K-Chips Act, passed in February 2025, increased tax credits for large semiconductor facility investments from 15% to 20% and for SMEs from 25% to 30% [19]. A bipartisan Special Semiconductor Act advanced to allow direct state subsidies and fast-track permitting for chipmakers [20].

Taiwan benefits from TSMC's dominance — the company controls roughly 64% of global foundry revenue and over 90% of advanced chip production at leading-edge nodes [7]. Taiwan's semiconductor market stood at $35.55 billion in 2025 and is projected to reach $51.88 billion by 2030 [21].

The concentration of public investment in chips raises a question: what about the rest of the industrial base? Steel, shipbuilding, petrochemicals, batteries, machine tools, and consumer electronics receive a fraction of the policy attention, even though they employ far more people.

Jobs, Demographics, and the SME Squeeze

The employment picture varies by country but shares a common thread: younger workers are leaving manufacturing, and traditional sectors cannot attract replacements.

In South Korea, December 2024 saw 52,000 net job losses — the first on-year decline since February 2021. Workers in their 20s lost 124,000 jobs during 2024, while those in their 60s and older added 266,000, reflecting an aging workforce unable to regenerate [22]. Manufacturing specifically shed 6,000 jobs. The OECD cut Korea's 2026 growth forecast from 2.1% to 1.7%, the second-largest downward revision among G20 nations [23].

Source: World Bank Open Data

Data as of Dec 31, 2024CSV

In Japan, the labor crisis takes a different form: not mass layoffs but a shortage of workers willing to enter manufacturing. A record 350 businesses went bankrupt due to labor shortages in 2024 [5]. Japan's SMEs, which form the supplier base for major manufacturers, are particularly exposed. The World Economic Forum noted that small businesses are receiving recruitment and retention support, but the structural decline in working-age population limits what policy can achieve [5].

Source: World Bank Open Data

Data as of Dec 31, 2025CSV

Taiwan lost about 4,000 manufacturing jobs in September 2025, with underemployment rising as output in traditional manufacturing slowed [24]. The proportion of Taiwan's employed population in manufacturing continues to decline while services employment grows [13].

Is This Actually Healthy? The Value-Chain Argument

There is a credible case that shedding low-margin commodity manufacturing is the correct economic trajectory for mature industrial economies. Japan, South Korea, and Taiwan all sit at the top of critical global supply chains in semiconductors, advanced materials, and precision equipment. Taiwan's economy grew 9% in 2025 precisely because it dominates the world's most valuable manufacturing niche [7].

South Korea's semiconductor output surged 13.2% in 2025 amid the AI boom, even as nearly every other sector contracted [2]. Japan's $65 billion semiconductor bet is designed to rebuild capacity in an industry it once dominated [14]. Patent data and R&D spending suggest these economies are investing in innovation: Japan's semiconductor strategy has attracted TSMC, Micron, Rapidus, and Kioxia/Western Digital to build or expand fabs [15][16].

The counterargument is that semiconductor gains benefit a narrow slice of the economy. TSMC's advanced fabs employ relatively few workers per dollar of output. The machine toolmaker whose exports fell 30% employs people in regional towns where there is no AI boom [8]. If the transition from commodity manufacturing to high-value services and advanced tech leaves behind entire regions and demographic groups — as it did in the U.S. Midwest and U.K. North — then the aggregate GDP numbers will mask real social costs.

Downstream Effects: SMEs and Supplier Networks

Japan's manufacturing sector depends on a vast network of small and medium enterprises that supply components to anchor industries. With 350 bankruptcies in 2024 attributed to labor shortages alone [5], and export competitiveness eroding under a weaker yen-dollar dynamic and tariff pressure, these supplier networks face compounding stress.

South Korea's 33 trillion won semiconductor package [18] will flow primarily to Samsung and SK Hynix, but the thousands of SMEs in petrochemicals, steel fabrication, and auto parts that form Korea's industrial midsection have fewer buffers. With domestic demand in a three-year decline and exports outside semiconductors shrinking, credit conditions for smaller manufacturers are tightening [6].

In Taiwan, the divide between the semiconductor ecosystem — centered around TSMC and its suppliers in Hsinchu — and traditional manufacturers elsewhere on the island grows wider with each quarter. The metals, plastics, and machine tools sectors that contracted through 2025 form the core of Taiwan's SME economy outside the tech corridor [8][9].

What Happens Next: Projections and Geopolitical Stakes

If current trends continue, the projections are sobering. The OECD forecasts South Korea's growth at just 1.7% in 2026, constrained by energy dependence, weak domestic demand, and export headwinds [23]. Japan's GDP grew only 0.1% in 2024, and the third-quarter 2025 contraction signals continued fragility [4].

Source: World Bank Open Data

Data as of Dec 31, 2024CSV

Taiwan's position is paradoxically the strongest and the most precarious. TSMC's dominance gives the island extraordinary geopolitical importance — Bloomberg Economics estimates a Taiwan conflict could cost $10 trillion globally, roughly 10% of world GDP [7]. But that same concentration creates systemic risk. China's PLA is reportedly targeting 2027 for Taiwan-scenario military readiness, and even short of conflict, a "quarantine" of Taiwan's shipping lanes could choke semiconductor exports without touching a single fab [7].

South Korea's ability to fund defense commitments — already strained by political turmoil in late 2024 and early 2025 — depends on sustained export revenues. If the margin compression in steel, petrochemicals, and batteries continues, fiscal space narrows [23][6].

Japan's strategic bet on semiconductor reshoring through Rapidus and TSMC's Kumamoto fabs will take years to mature. Rapidus aims for 2nm mass production by 2027 — a timeline many analysts view as ambitious given that only TSMC and Samsung have achieved comparable nodes [16]. In the meantime, the broader manufacturing base continues to shrink under demographic and competitive pressure.

The Road Ahead

The industrial contraction facing Japan, South Korea, and Taiwan is neither a sudden crisis nor an inevitable decline. It is a structural transition accelerated by Chinese overcapacity, demographic aging, and trade fragmentation — but one that also reflects the natural evolution of high-income economies toward services and advanced technology.

The critical variable is speed. The U.S. and U.K. deindustrialized over decades, and the social and political costs — regional inequality, populist backlash, hollowed-out communities — are still being paid forty years later [10]. East Asia's transition is happening faster and under greater geopolitical pressure. Whether these three economies can manage the shift without leaving behind the workers and regions that built their prosperity will depend on whether industrial policy extends beyond semiconductors to address the full breadth of the manufacturing base now under strain.