South Korea navigated the 1997 Asian crisis with an IMF dictating reforms and a plummeting won. It weathered 2008 with a banking sector exposed to global turbulence. In both cases, it rebounded in less than two years. What has been happening since 2023 is different: very weak growth (~1.4-1.6%), pulled down by well-identified external shocks — collapse of semiconductor exports, global slowdown, monetary tightening — but without an open financial crisis. The breakdown is also coming from within.

This slowdown is occurring in a country that remains one of the world’s most efficient industrial machines. Industry represented 31.6% of South Korean GDP in 2023 according to the World Bank, compared to 22% on average in OECD countries and approximately 13-13.4% in France according to Eurostat data. Sixty years of disciplined upgrading have produced an industrial apparatus without equivalent in the category of high-income countries. The question is not whether this model has succeeded. It is why it is no longer sufficient, and what can take its place.

The Essentials

  • Industry represents 31.6% of South Korean GDP in 2023, versus 22% average across OECD, according to the World Bank.
  • South Korea has experienced very weak growth from 2023 onward, linked to identified external shocks (collapse of semiconductor exports, global slowdown, monetary tightening), a situation unprecedented even compared to the crises of 1997 and 2008.
  • The fertility rate fell to 0.72 children per woman in 2023, the lowest in the world, implying a reduction in the active workforce of approximately 40% by 2065 according to Statistics Korea projections.
  • The chaebols (Samsung, Hyundai, SK, LG) represent a significant share of national exports, creating structural dependence on a few sectors exposed to semiconductor cycles.
  • The government has committed to a plan supporting natality exceeding $200 billion cumulatively since 2006, without demonstrable results at this stage on demographic indicators.

Sixty Years of Upgrading Without Precedent

The model must be explained before its breakdown can be diagnosed. In 1960, South Korea’s per capita income was lower than Ghana’s. Fifty years later, the country joined the OECD as a full member, with an industrial base covering shipbuilding, consumer electronics, semiconductors, petrochemicals, automobiles, steel — and an agriculture that underwent its own green revolution, little known abroad but documented by statistics from the Korean Ministry of Agriculture.

This trajectory is not explained solely by the chaebols, those family conglomerates that economic newspapers describe in every article about South Korea. It is explained by a rare combination: a constant priority given to technical and university education, sustained public investment in R&D (the country devotes approximately 4.9% of its GDP to research and development according to the OECD, placing it third globally), and a culture of continuous improvement that has permeated sectors well beyond electronics. Samsung is the visible head; the model is in the roots.

What this model has accomplished is nothing automatic. It results from precise political choices: the deliberate industrial policy of the Park Chung-hee years, the brutal reform of the financial system after 1997, the massive requalification of the workforce in the 2000s. Public and private actors took bets, often risky, often costly in the short term. The OECD has documented how public spending on education and continuous training accompanied every major sectoral transition. This is not a lesson to be imported mechanically, but it deserves to be read in its entirety before being judged.

Why Current Stagnation Resembles No Other

The crises of 1997 and 2008 had external triggers: currency collapse for the first, global credit paralysis for the second. Response policies were identifiable and measurable. South Korea was able, in both cases, to mobilize its institutions and enterprises to absorb the shock and rebound.

The current stagnation is of a different nature. It stems from three simultaneous factors that reinforce each other mutually. The first is cyclical: the semiconductor cycle underwent a pronounced trough in 2022-2023, penalizing Samsung and SK Hynix, which alone weigh a determinant share of national exports. The second is structural: domestic demand is constrained by household debt that has reached levels among the highest in developed countries, according to the Bank for International Settlements. The third is geopolitical: Korean trade rests on an export architecture toward both China and the United States simultaneously, and trade tensions between the two powers have placed Seoul in an uncomfortable balancing act, with no good option.

What distinguishes this episode from previous ones is the absence of visible recovery mechanisms. In 1998, the IMF’s shock therapy was harsh but its effect was readable: reduce costs, restructure banks, allow nonviable firms to disappear. Today, the classic levers are not working. The Bank of Korea has lowered its rates. The government has announced support plans. Chip exports have resumed since late 2023 with AI demand. Yet growth remains sluggish, domestic consumption is not recovering, and business confidence according to indices from the Korea Development Bank is stagnating.

The Demographic Wall That Changes the Nature of the Problem

The real long-term issue is not cyclical. It is demographic. In 2023, South Korea recorded a fertility rate of 0.72 children per woman, the lowest ever measured in a developed country. For reference, the generational replacement threshold is 2.1. Japan, long cited as an extreme case of aging, hovers around 1.2.

Statistics Korea projections, the national statistical agency, place the reduction in the active workforce at approximately 40% by 2065, under the central hypothesis that assumes a slight increase in fertility toward 1.0 over the coming decade. If the current trend continues without correction, the horizon narrows further. This is not a futuristic projection: it is the mechanical consequence of a population pyramid already drawn. The children who will enter the labor market in 2040 are already born, or will not be.

The South Korean government has spent, since 2006, more than $200 billion on natality support policies according to figures compiled by the Ministry of Health and Social Affairs: allowances, parental leave, subsidized childcare. The result is null on fertility indicators. This does not mean these policies are useless — they have improved living conditions for families with children. But they have not reversed the trend, because the underlying causes are not financial. They concern the cost of education, which weighs on Korean households far beyond public schooling, the intense competition for access to universities and major corporations, and a housing market in major cities that makes life with two children economically exhausting for young couples.

Where other countries have used immigration to compensate for the natural decline in their active workforce, South Korea remains reluctant. The society remains more than 96% homogeneous according to national census data. The debate on immigration is not absent, but it is at the preliminary stage of a social conversation that will take time before producing policies.

This context sheds different light on the question of finance and resource allocation: in an economy where the workforce will mechanically contract, the question is no longer how to mobilize more workers, but how to produce more value with fewer people. This is a constraint that entirely redistributes investment priorities.

Current Bets: AI, Automation, and Chaebol Model Reform

South Korea is not without answers to this challenge. Several dynamics merit being described concretely.

The first is automation. South Korea is, by far, the most roboticized country in the world: according to the International Federation of Robotics, it had 1,012 industrial robots per 10,000 employees in manufacturing in 2022, far ahead of Singapore (730) and Germany (415). This is not the result of recent policy. It is the accumulation of industrial investments over thirty years. But the current acceleration, driven by AI and semiconductor needs, is pushing Korean companies to integrate increasingly sophisticated systems not only in heavy industry, but in services and logistics. The question of computing power as a scarce resource takes on concrete dimension here: the country that manufactures chips and has the highest robot density in the world is structurally positioned to profit from the AI transition — if productivity gains materialize at the economy-wide scale.

The second is the reorientation of the higher education system. Several Korean universities, including KAIST (Korea Advanced Institute of Science and Technology) and Pohang University, have formed partnerships with semiconductor and AI companies to create short curricula oriented toward specific skills rather than prestigious degrees. This is a cultural shift in a country where the rank of the university of entry still largely structures professional trajectories.

The third is more difficult to quantify: growing pressure for chaebol model reform. The great families controlling Samsung, LG, Hyundai, and SK have long benefited from a privileged relationship with the South Korean state, documented notably by the work of the Korea Fair Trade Commission on market concentration. Governance scandals of the 2010s (including the 2021 conviction of Samsung Vice President Lee Jae-yong) fueled public debate on the capture of model gains by a very limited number of families. The Korean Competition Commission has hardened its procedures. Korean tech and biotech startups have begun attracting independent financing. The issue is whether a decentralized innovation economy can emerge alongside — and not only below — the chaebols.

What the Korean Experience Tests on a Global Scale

South Korea is an extreme case, but not an isolated one. Japan experienced its own demographic stagnation and its own dependence on an industrial model built for extensive growth. Italy has the same structure of family businesses without the same industrial capacity. Germany combines aging and partial deindustrialization.

What South Korea is testing in real time is the possibility of moving from a catch-up model founded on discipline and volume to a productivity model founded on technological intensity, in a context of contracting human resources. It is a problem that growth economists have theorized but that few countries have had to solve as frontally, as quickly.

Philippe Aghion, whose work on Schumpeterian growth has precisely analyzed the transition between imitation and innovation, has documented how catch-up countries run out of steam at the technological frontier because institutions, incentives, and human capital structures remain optimized for imitation. South Korea has been at this frontier for more than ten years. Its ability to finance the next generation of its own R&D, to keep its engineers on its soil rather than let them depart for California campuses, and to create the conditions for an innovation capitalism less concentrated than the chaebol model, is what will determine the trajectory over the next twenty years.

The question of human capital is directly linked to the demographic challenge. When the active workforce shrinks 40% over forty years, the alternative to structural recession is a productivity increase per worker of a magnitude that few countries have achieved. OECD data show that South Korean labor productivity per hour, while increasing, remains below the average of Northern European countries — partly because the culture of long working hours has substituted volume for efficiency. Labor law reforms undertaken under several successive governments to reduce actual working time and improve conditions in SMEs have produced only partial results.

The Challenge of a Second Transformation

South Korea accomplished a first transformation in the early 2000s: moving from an imitation economy to an economy producing at the technological frontier. Samsung is no longer merely an assembler — it is one of the rare global actors capable of designing and manufacturing chips smaller than 3 nanometers. Hyundai designs its own electric vehicle architectures. KAIST produces researchers cited in Nature and Science.

The second transformation is of a different nature. It supposes building an economy that creates value with a shrinking workforce, in sectors where the Korean advantage is no longer cost but ingenuity, and within a social framework where younger generations refuse the working conditions that made the miracle possible. This is a less visible transformation than the first, because it affects institutions, norms, and incentives rather than factories and universities.

What makes the outcome uncertain is not the absence of resources or talent. It is the speed at which demographics constrain choices. A country with forty years to reform its institutions has real room for maneuver. A country whose active workforce is beginning to decline now has a narrower window to initiate the transition before contraction weighs on tax revenues, pensions, and public investment capacity.

The indicator to watch is not the quarterly growth rate. It is the participation rate of women in the South Korean labor market, which remains significantly below that of men according to OECD data, and which represents the most immediately mobilizable reserve of labor. It is also the level of R&D investment by non-chaebol firms, which would measure the emergence of a less concentrated innovative fabric. And it is, perhaps above all, the capacity of the South Korean education system to form young people capable of creating companies rather than joining large ones — a cultural reversal that neither subsidies nor curriculum reforms have yet managed to produce, but which some emerging signals in Seoul’s startup ecosystem allow one to envisage.

Sources

  1. General Directorate of the Treasury — Indicators and Situation South Korea: https://www.tresor.economie.gouv.fr/Pays/KR/indicateurs-et-conjoncture
  2. World Bank — Industry Data (% of GDP): https://data.worldbank.org/indicator/NV.IND.TOTL.ZS
  3. OECD — R&D Spending as % of GDP: https://data.oecd.org/rd/gross-domestic-spending-on-r-d.htm
  4. Statistics Korea — Demographic Projections: https://www.kostat.go.kr/
  5. International Federation of Robotics — World Robotics Report 2023: https://ifr.org/world-robotics
  6. OECD — Labor Productivity and Employment Rates: https://stats.oecd.org/
  7. Bank for International Settlements — Household Debt: https://www.bis.org/statistics/totcredit.htm
  8. World Bank — Industry South Korea 2023: https://tradingeconomics.com/south-korea/industry-value-added-percent-of-gdp-wb-data.html
  9. World Bank / Statistics Korea — TFR South Korea 2023: https://tradingeconomics.com/south-korea/fertility-rate-total-births-per-woman-wb-data.html
  10. IFR — Robot Density South Korea 2022: https://ifr.org/ifr-press-releases/news/global-robotics-race-korea-singapore-and-germany-in-the-lead
  11. OECD MSTI / SSTI — R&D South Korea 2023: https://ssti.org/blog/useful-stats-international-comparison-rd-expenditures
  12. BIS / Asia Business Daily — Household Debt South Korea: https://www.asiae.co.kr/en/article/2026011514373334962
  13. IFRAP / Eurostat — Industrial Share France 2023: https://www.ifrap.org/emploi-et-politiques-sociales/il-manque-241-milliards-la-valeur-ajoutee-industrielle-de-la-france-par-rapport-lallemagne
  14. Statistics Korea / Morgan Stanley — Active Population 2065: https://www.morganstanley.com/ideas/south-korea-population-decline-aging-crisis
  15. Newsweek / CNN — Natality Plan +$200B since 2006: https://www.newsweek.com/south-korea-news-plans-tackle-population-crisis-2025-2007358