Can Korea Turn a Semiconductor Tax Boom Into the Next Growth Cycle?

South Korea wants to turn a semiconductor-driven tax windfall into infrastructure, skills and regional growth. The harder challenge is making sure the resulting productivity reaches firms, workers and cities beyond the technology companies already benefiting from the boom.

South Korea entered 2026 expecting ₩390.2 trillion in national tax revenue. By the end of March, the government had raised that estimate to ₩415.4 trillion, an increase of ₩25.2 trillion that was almost exactly the size of the supplementary budget submitted to the National Assembly at the same time. The National Assembly Budget Office produced a slightly lower revenue estimate of ₩413.8 trillion, yet the direction was unmistakable: after two years dominated by revenue shortfalls, the public finances had swung abruptly toward abundance.

Corporate earnings supplied much of the improvement. Semiconductor profits had recovered sharply, equity markets were stronger and related income was feeding into several tax categories. The additional fiscal room arrived just as the government faced an energy shock and weaker conditions elsewhere in the domestic economy, so the supplementary budget directed new spending toward fuel-price relief, vulnerable households, affected industries and supply-chain stabilization. Korea’s existing budget system had done what it was designed to do: revenue discovered during the fiscal year became available for the pressures confronting the government in the same year.

The proposed Future Response Fund would send part of the next revenue surge through a different channel. Legislation placed on public notice on Aug. 24 would establish the fund to reduce instability from tax fluctuations while financing investments intended to raise Korea’s potential growth rate. Government plans identify youth, future growth industries, regional development and education and talent as the principal areas for investment, with accumulated resources also available when fiscal conditions weaken. The bill remains a proposal and its eventual scale will depend on the legislative process and future tax forecasts, but the economic ambition is already clear.

The central government is trying to separate part of the income produced during an unusually favorable cycle from the spending demands of the year in which it arrives. Semiconductor profits can fall quickly; a transmission network, research capability, a skilled workforce or a regional industrial base can continue generating income long after memory prices have turned.

KDI’s August outlook gives the strategy an unusually sharp economic setting. The institute expects GDP to grow 3.2 percent in 2026, with exports rising 8.7 percent and equipment investment 7.9 percent as the global semiconductor and AI investment cycle remains exceptionally strong. Private consumption is expected to increase by a more modest 2.3 percent because income gains remain concentrated in the semiconductor sector. KDI expects growth to slow to 2.2 percent in 2027 even as semiconductor-related investment remains strong.

Korea can therefore produce extraordinary export earnings, corporate profits and tax receipts without spreading the same improvement through household demand, smaller companies or regional economies at the same speed. The proposed fund matters because it attempts to use the fiscal system to bridge that gap. Its value will depend far less on the amount of money accumulated than on whether exceptional revenue leaves behind sources of income that no longer depend on the exceptional year that produced it.

The Revenue Behind the Boom

Korea does not collect a semiconductor tax. Corporate tax, income tax, securities-related taxes and consumption taxes enter the Treasury through separate legal channels, even when part of the economic impulse behind them originates in the same industrial cycle.

Strong memory prices and AI-server demand raise profits at semiconductor manufacturers and suppliers. Higher profits can support bonuses, investment and equity valuations; compensation produces additional income-tax receipts, financial-market activity raises transaction-related revenue, and supplier and household income can later feed into consumption taxes. The fiscal accounts gradually show a broader improvement even when a meaningful share of the originating shock remains concentrated in a relatively small group of globally competitive businesses.

Recent corporate-tax history shows how quickly those conditions can reverse. Receipts fell from ₩103.6 trillion in 2022 to ₩80.4 trillion in 2023 and ₩62.5 trillion in 2024 as corporate earnings weakened. Internal-tax revenue followed the cycle in less dramatic but still consequential fashion, declining from ₩352.3 trillion in 2022 to ₩306.1 trillion in 2023 and ₩298.3 trillion in 2024. The swings were large enough to force substantial adjustments in central and local public finances.

Tax receipts also arrive with a lag. Profits are earned before they are fully reflected in annual corporate-tax settlements, while bonuses, equity transactions and consumption propagate the original income shock on different schedules. A government can therefore begin allocating the fiscal proceeds of a boom after the underlying industrial cycle has already moved forward.

The proposed fund attempts to change the useful life of that income. Revenue generated by a short upswing could finance assets whose benefits extend across several business cycles: power infrastructure serving multiple companies, common research or testing equipment, supplier capabilities, university-industry systems and workers whose skills remain valuable after the original source of tax revenue weakens.

Government material published on Aug. 25 defines additional tax revenue as the portion of the next year’s internal-tax budget that exceeds a long-term trend, while excess tax revenue arises when a later revenue re-estimate exceeds the internal-tax budget already in place. The government has said the actual amount available to the fund will be announced with the 2027 budget and the national fiscal management plan.

The trend calculation deserves attention because legislation would turn a statistical estimate into a fiscal boundary. Applying different reasonable ways of summarizing the same historical tax series can place tens of trillions of won on different sides of that boundary. A reconstruction using an endpoint growth rate and a simple log-linear fit, for example, produces a difference of roughly ₩37 trillion in one recent year and more than ₩40 trillion in the next. The exercise does not establish a superior formula. It shows how strongly the classification of temporary revenue can depend on the method chosen once tax history contains both an extraordinary boom and a sharp reversal. The underlying historical series itself shows internal-tax revenue moving from ₩182.0 trillion in 2015 to ₩352.3 trillion in 2022 before falling below ₩300 trillion in 2024.

Local governments will feel the consequences of that classification immediately. The Local Allocation Tax amendment released on Aug. 24 would remove amounts transferred to the Future Response Fund from the internal-tax base used to calculate the allocation, while creating authority to support local governments through the fund when tax conditions deteriorate. Korea is therefore proposing to replace part of local governments’ automatic participation in an unusually strong revenue year with a system in which the central government pools more of the upside and assumes a larger role when the cycle reverses.

Under the existing 19.24 percent statutory share, every ₩10 trillion removed from the relevant internal-tax base corresponds to ₩1.924 trillion that no longer enters the initial Local Allocation Tax calculation. The amount should not be read as a net loss to local governments because national investment and later support can return resources through other channels. The shift changes who controls the money first, when it can be spent and how the fiscal consequences of the next downturn are shared.

A well-designed revenue rule can stabilize the entrance to the system. Economic value is created only after the money leaves the fund.

What Public Money Can Add

Korea’s largest semiconductor companies do not suffer from a general lack of access to capital. Their global earnings, balance sheets and access to financial markets allow them to finance fabrication plants, equipment and research on a scale that dwarfs most individual government programs. Public support for those investments may serve strategic purposes, particularly when global competitors offer large incentives of their own, yet the headline size of a subsidized project reveals little about how much investment the state actually caused.

Electricity provides a clearer case for public intervention. Semiconductor clusters and AI data centers need generation, transmission and grid connections that cross jurisdictions and serve multiple users over decades. No single company can capture all of the benefits from expanding the network, while delays can prevent several private projects from proceeding at once. Water systems, transport links, common testing facilities and research infrastructure create similar coordination problems.

Government spending has greater leverage when it removes such shared constraints. A transmission line that brings forward several private investments, a testing platform that allows smaller suppliers to qualify for international customers or a university research facility that provides equipment beyond the reach of an individual regional company can generate economic activity far larger than the original appropriation.

Korea’s productivity structure makes that leverage especially important. OECD analysis has documented persistent productivity gaps between large Korean corporations and smaller companies, with many SMEs concentrated in lower-productivity activities even inside an economy that contains some of the world’s most technologically advanced manufacturers. An industrial boom centered on a handful of frontier firms does not automatically raise the performance of the firms around them.

Supplier policy can change that relationship when public money leaves a company able to do something it could not do before. Access to advanced packaging, testing, industrial AI, certification, shared equipment or research staff can allow a manufacturer to move into more demanding contracts and new export markets. Temporary orders financed by subsidy can raise sales for a year without changing the supplier’s technology, customers or ability to finance its next investment.

Workers face a comparable divide. Korea has the highest tertiary-attainment rate among young adults in the OECD: 71 percent of 25- to 34-year-olds hold a tertiary qualification, compared with an OECD average of 48 percent. Employment among tertiary-educated young adults is only 80 percent, below the OECD average of 87 percent. Separate OECD adult-skills data find that 31 percent of Korean workers are over-qualified for their jobs and 49 percent work in fields that do not closely match their qualification.

Another large training target does little by itself to resolve those mismatches. Public spending becomes more valuable when participants move into work where the new skills are used, earn more and remain in those occupations after the program ends. Training tied to employers, equipment, apprenticeships and regional research projects is more likely to change a worker’s economic position than a program measured primarily through enrollment and course completion.

Housing and transport around new industrial centers can belong to the same investment strategy. A cluster can have vacancies for highly skilled workers and still fail to build a durable labor market if housing prices, commuting or weak local services discourage people from settling near the new jobs. Industrial infrastructure ends at the factory gate only on a construction plan; the labor market required to operate the factory extends into the surrounding city.

Public money drawn from a temporary tax surge will create the greatest long-term value where it allows firms and workers to make investments or transitions that private markets and existing institutions were failing to support. Subsidizing an activity that was already profitable may still have a policy rationale, but it should not be confused with creating new productive activity.

How Far the Productivity Travels

KDI’s forecast already reveals the central weakness in relying on the semiconductor boom to spread by itself. Exports and equipment investment are rising rapidly, while household consumption improves much more slowly because income growth remains concentrated in the leading sector. Public investment can broaden the boom only if new capabilities begin to appear outside the companies already generating the strongest profits.

Smaller manufacturers offer the most direct route. Suppliers that acquire new technology, enter higher-value stages of production or begin selling to customers beyond a single dominant buyer can retain part of the productivity created around semiconductor and AI investment. Higher value added then appears in a wider range of wages, profits, purchases and future investment rather than remaining inside a small group of corporate balance sheets.

Labor mobility determines how widely workers share in the same process. Technology booms initially raise the return to scarce engineering and technical skills, rewarding people who already possess them. Public investment widens participation when workers outside established technology careers can finance training, relocation and periods of transition and then move into jobs where productivity and earnings are higher.

Regional economies present the most difficult version of the problem because fixed assets are easier to relocate than the networks that determine where value is retained. OECD research finds that productivity in the Seoul Capital Area is now about 25 percent higher than in Korea’s other large metropolitan regions, while those large cities sit roughly 23 percent below the national average. Between 2008 and 2022, productivity in the capital region grew at an average annual rate of 1.66 percent compared with 1.49 percent nationally.

A fabrication plant, AI data center or national research institute outside Seoul changes the map of investment immediately. High-value engineering contracts can still go to companies headquartered elsewhere; specialized equipment can be imported; senior workers can remain tied to the capital region; profits can accrue to parent companies whose financial and professional networks stay in Seoul. A region can host billions of won in physical assets while retaining a much smaller share of the knowledge, wages and business capability generated around them.

Local supplier networks, university research, technical-service companies and permanent skilled employment determine whether the physical investment develops into a regional economy. Data centers illustrate the problem particularly well. Servers, cooling systems and power infrastructure can create a very large capital stock with comparatively few permanent jobs. A region captures much more when computing capacity supports nearby software businesses, manufacturers, researchers and energy-service companies rather than functioning mainly as infrastructure serving users elsewhere.

Busan provides a useful baseline because its own medium-term fiscal plan already shows the difference between receiving public money and generating more revenue locally. The city projects own-source revenue to rise from ₩6.77 trillion in 2026 to ₩7.17 trillion in 2030, an increase of about ₩401 billion. Transfer revenue is projected to rise from ₩9.44 trillion to ₩11.02 trillion, adding roughly ₩1.57 trillion over the same period. Transfers are therefore expected to grow nearly four times as much as own-source revenue.

National transfers remain essential to Busan’s finances and to fiscal equalization across Korea. Regional industrial policy promises something more than a larger transfer. Investments in ports, logistics, advanced manufacturing, AI or research would have altered the city’s economic base if a later fiscal plan begins to show faster growth in locally generated revenue because firms, wages and taxable business activity have expanded.

A city gains more than an asset when the income created around that asset remains after the national appropriation has ended.

Why Inequality Belongs Inside the Growth Story

The Bank of Korea’s June analysis of household polarization shows how concentrated growth can eventually become a productivity problem of its own. Korea’s net-asset Gini coefficient rose from 0.584 in 2017 to 0.625 in 2025, while the disposable-income Gini edged upward from 0.323 in 2023 to 0.325 in 2024 after years of improvement. Among households simultaneously in the lowest fifth for income and net assets, the share headed by people in their 20s and 30s increased from 7.9 percent in 2020 to 15.2 percent in 2025.

The Bank’s cross-country analysis also found that a one-percentage-point increase in the share of assets held by the richest 10 percent was associated with a 0.16 percent decline in total-factor productivity. The estimate does not imply that every redistribution policy will raise productivity, and the relationship should not be read as a simple one-way causal law. It does indicate that concentration can reduce the efficiency with which people and capital reach their most productive uses.

Household wealth affects who can afford to spend time acquiring new skills, relocate to an emerging industrial center or enter a housing market near high-paying work. Financial constraints can keep capable workers in occupations below their potential and limit the ability of smaller entrepreneurs to finance productive opportunities. Regional disparities create a parallel barrier when advanced jobs, research institutions and professional networks remain concentrated in places that are expensive or difficult for outsiders to enter.

Youth and regional programs financed from the proposed fund therefore have an economic role when they widen access to productive work. Affordable housing around a new cluster can matter when rents would otherwise prevent younger workers from moving there. A regional university becomes part of industrial development when its laboratories and curricula connect students and local firms to the technologies being deployed nearby. Supplier finance becomes valuable when it allows a viable company to acquire equipment or certification that commercial lenders will not finance on workable terms.

Korea already produces more university graduates than its labor market uses efficiently. Another cohort of trainees will add little if the most valuable opportunities remain concentrated within the same companies, universities, neighborhoods and professional networks. Spending aimed at narrowing gaps becomes part of a growth policy when people who previously stood outside high-productivity sectors begin generating higher-value income themselves.

Public policy still has legitimate welfare purposes that need no productivity justification. Income support can be warranted because households need protection, and public services do not have to produce a measurable return on GDP to deserve funding. Temporary revenue explicitly set aside for future-oriented investment, however, should preserve a visible connection between what is spent today and the economic capabilities expected to remain tomorrow.

The same discipline applies to regional allocations. Money can be assigned outside Seoul immediately; local business capability, wages and research networks take years to build. The economic significance of the spending appears after construction ends, when firms continue operating, workers remain and the region can sustain more activity without another appropriation of the same kind.

Ireland’s Warning

Ireland has already lived through a version of Korea’s fiscal dilemma. Corporation-tax receipts generated by a small group of highly profitable multinational companies became large enough to transform the public finances, creating headline surpluses while exposing the budget to decisions and profits concentrated in a narrow corporate base.

The Irish government responded by separating some of that fiscal strength into two funds with different purposes. The Future Ireland Fund receives an annual contribution linked to GDP for long-term expenditure pressures, while the Infrastructure, Climate and Nature Fund is designed to support capital expenditure through weaker economic periods and designated environmental investment. Ireland had transferred €10.4 billion into the two funds by the end of 2024.

Separate accounts solve only part of the problem. The Irish Fiscal Advisory Council estimated in June that Ireland would run an underlying deficit of about €11 billion in 2026 after excess corporation-tax revenue was removed, equivalent to roughly 3 percent of modified gross national income. Under the government’s plans, the Council calculates that only about €1 of every €6 collected in corporation tax will be saved, with the remainder supporting current commitments. It also warns that some future payments into the savings funds may have to be financed through borrowing.

Ireland can therefore accumulate substantial financial assets while the ordinary budget becomes increasingly accustomed to revenue that its own fiscal watchdog considers unusually risky. Korea could face the same contradiction if several consecutive years of semiconductor-driven receipts raise the spending baseline outside the Future Response Fund.

A useful Korean fiscal account would show the budget balance after revenue classified as additional or excess has been removed. Persistent ordinary expenditure should remain supportable by revenue that policymakers consider durable. A growing fund alongside a worsening balance on that basis would show that the windfall had been separated institutionally while continuing to finance the state indirectly.

Norway offers a narrower lesson in separating the period when exceptional income is earned from the period when it is spent, while Chile’s structural-balance system shows why a revenue benchmark with large fiscal consequences benefits from continuing independent scrutiny. Korea has little reason to copy either model wholesale. Semiconductor production is an industrial capability the country wants to expand, rather than a finite natural resource whose proceeds can simply be invested abroad.

Korea’s harder problem comes from asking a single fund to perform several functions at once. Resources may be needed for a future revenue downturn, while strategic infrastructure can lose economic value if investment is postponed for years. Regional programs, industrial projects and human-capital commitments also create different time horizons and future obligations.

A headline fund balance will become increasingly uninformative as those commitments accumulate. A fund holding ₩100 trillion in assets may have far less than ₩100 trillion readily available during a downturn if much of the money has already been contracted to power systems, research projects or multi-year programs. Staffing, maintenance and operating costs can continue after the original investment has been exhausted, transferring obligations created during a boom into later ordinary budgets.

Public accounts will eventually need to distinguish liquid resources from committed investment and show the recurring costs created by projects financed with temporary revenue. The same won cannot remain fully available for a future downturn after the government has effectively promised it to a long-lived project.

Ireland’s experience makes the broader risk clear: governments can save a visible portion of a windfall while quietly allowing the rest of the fiscal system to behave as though the exceptional income will continue.

What Remains After the Boom

By the end of the decade, Korea will be able to count the money deposited in the Future Response Fund, the projects approved, the workers trained and the amount allocated outside the capital region. Those figures will describe the scale of government activity. The more consequential evidence will appear in the economy after the original support recedes.

A semiconductor supplier that used public infrastructure to enter higher-value production should eventually be able to invest from its own earnings and sell to customers it could not previously reach. A worker who received advanced-industry training should still be earning more in productive employment several years later. Private investment should appear around publicly financed grids, research platforms and transport links because those assets changed the economics of projects that firms had previously delayed or rejected.

Regional cities should also begin retaining more of the income generated around national investment. New technical jobs, suppliers, university-industry links and business formation would matter more than the construction value of the original project. Busan could continue receiving substantial national transfers while simultaneously generating more revenue from its own expanding economy; regional development does not require the disappearance of fiscal equalization, only a stronger productive base beneath it.

The national budget will have its own reckoning. Permanent spending should remain financeable when exceptional tax revenue is removed, while enough uncommitted resources remain available to absorb the next serious revenue decline. Long-lived infrastructure can justify long-lived operating costs when it raises productivity and future tax receipts, but those obligations should be visible before the temporary revenue that initiated them has gone.

Semiconductor cycles will continue to turn. KDI already expects the extraordinary conditions of 2026 to moderate in 2027 even while the Korean economy remains relatively strong. Memory prices, AI investment, global competitors and technology itself will move in ways that no fiscal rule can forecast with precision.

The proposed Future Response Fund cannot eliminate those cycles, and Korea has little reason to try. A more durable achievement would be to prevent an unusually profitable period from disappearing entirely into expenditure whose economic effect ends with the budget year.

If the next semiconductor downturn finds smaller manufacturers producing more valuable goods, workers outside the existing technology elite earning more, regional cities retaining a larger share of advanced economic activity and the ordinary budget able to function without another exceptional tax year, temporary revenue will have changed the structure from which future taxes are collected.

Korea will know whether that happened only after the current boom has ceased to look exceptional. By then, the important number will no longer be the amount of tax revenue the semiconductor cycle once produced, but the income, productive assets and broader tax base that remained after it had moved on.