South Korea's government will treat the IMF's upgrade as a victory lap, and the markets will cheer. A 3.9% growth forecast is a handsome number. But look at what is actually producing it: record chip exports, driven by a global scramble to build artificial intelligence infrastructure. That is a boom in a few enormously profitable companies, not a rising tide.
The IMF is polite about this, but the admission is in the report: the benefits are spreading only slowly beyond the technology sector. Translate that from fund-speak. A bakery owner in Gwangju, a part-time worker in a convenience store, a 70-year-old paper collector in Daegu are not feeling a 3.9% economy. Korea still has the highest elderly poverty rate in the OECD, a labor market split between protected regular workers at big firms and precarious workers at subcontractors, and a capital region that keeps swallowing the young and the money. A semiconductor windfall does nothing about any of that unless politics makes it.
Here is the question the headline buries: who owns the windfall? Korea's export machine was built with decades of public sacrifice: cheap state credit, suppressed wages, tax breaks, and ordinary citizens bailing out chaebol-linked firms in 1997. When a handful of firms now earn historic profits from a global bubble of demand they did not create, the public has a legitimate claim on a share. Some will say that is resentment. It is arithmetic. A capital-intensive AI boom creates relatively few jobs for the amount of money it generates. If the gains sit in corporate balance sheets, dividends and a rising stock market, the country gets richer on paper while most households get a higher cost of living.
The serious objection is that Korea is in a brutal technology race with China, Taiwan and the United States, and that squeezing the golden goose invites the goose to relocate. Fair. Nobody serious proposes confiscating investment capital. But there is a wide gap between wrecking competitiveness and demanding that the winners of the boom fund the things that make it durable: regional infrastructure outside Seoul, a real pension floor for the elderly, childcare and housing that let young workers consider having children in a country with one of the lowest birth rates ever recorded, and wage-setting that reaches subcontractors rather than stopping at the headquarters gate.
The IMF's prescription deserves scrutiny too. "Contain demand pressures" while offering "targeted support to vulnerable households" is the familiar formula: keep the checks small, narrow and means-tested, and tighten when the numbers look hot. But the pressure on Korean households is not too much demand. It is housing costs, stagnant pay outside the export core, and old-age insecurity. Targeted crumbs are a poor answer to a structural gap.
The Left in Korea has its own test. The Democratic Party has made labor-friendly gestures and talks about shared growth, and it will be tempted to bask in a strong GDP print and a soaring stock index. That would be a failure. If the government that promised to rebalance power between big capital and everyone else uses an AI-export boom as an excuse to postpone the harder work, it will have squandered the one moment when the money is actually there.
Booms end. Chip cycles are among the most violent in industry. The time to build a fairer economy is while the profits are record-setting, not after the downturn, when the same voices will say there is nothing to share. Korea should celebrate the number for about a day, then ask who gets the money.
How it may affect me
A strong headline forecast does not guarantee gains for most households. With inflation projected at 2.7% in 2026, wage growth outside the chip and export sectors may not keep up, especially for subcontract and part-time workers. Watch for whether the government's support is broad or narrowly means-tested, since that will decide who actually benefits.


