For forty years, the world shipped its manufacturing to wherever labour was cheapest. Then three shocks, in three years, changed the calculation permanently. The reindustrialisation wave — America's CHIPS Act and IRA, Europe's Critical Raw Materials Act, India's PLI schemes — is the largest peacetime industrial policy in human history. The factories are being built. The capital is committed. 244,000 manufacturing jobs reshored to the US in 2024 alone, pushing cumulative reshoring past 2 million since 2010. But the bottleneck is not capital or policy or will. It is workforce. This letter is about who is genuinely winning and who is kidding themselves.
Not investment advice. Data sourced from Reshoring Initiative 2024 Annual Report, Capgemini Research Institute Reindustrialisation Study Edition 3 January 2026, SEMI Europe Chips Act Report November 2025, SAIS Review US-India Semiconductor Partnership, EU Critical Raw Materials Act RESourceEU December 2025. All figures current as of June 2026.
Countries with empty factories tried to buy ventilators, masks, and PPE on global spot markets. Countries with the factories kept them. The US imported 90% of its pharmaceutical ingredients from China and India. When supply chains froze, hospital systems ran out of basic equipment. The lesson: just-in-time manufacturing at global scale is efficient in stable conditions and catastrophically fragile in disrupted ones.
Europe had built its entire industrial model on cheap Russian natural gas. When the invasion made that gas unavailable or politically unacceptable, German industrial competitiveness collapsed overnight. Energy prices quintupled. BASF moved production to the US. Volkswagen discussed factory closures. The lesson: a manufacturing base dependent on a single supplier for its critical input is not a manufacturing base — it is a liability waiting to be called.
TSMC produces more than 90% of the world's most advanced semiconductors on an island that China claims as its own territory. Economists estimate a Chinese blockade scenario would inflict $2.7 trillion in global economic losses within the first year. The car industry got a preview in 2021 when a chip shortage stopped production lines across the world. The lesson: the most critical input in the modern economy is almost entirely made in one place.
Three shocks. Three different inputs. Three identical conclusions. The globalisation model — make things wherever labour is cheapest, trust supply chains to deliver reliably, assume geopolitics won't interrupt commerce — is structurally inadequate for a world of heightened geopolitical tension, climate-driven supply disruption, and technological competition. The reindustrialisation wave is not protectionism for its own sake. It is the rational response to the discovery that efficiency and resilience are different objectives — and that the world spent forty years optimising for efficiency at the cost of resilience. The bill arrived between 2020 and 2022.
The CHIPS and Science Act (2022) committed $52 billion to domestic semiconductor manufacturing — the explicit acknowledgement that a country cannot be a technology superpower without a domestic chip-making capacity. To date, over fifty new semiconductor ecosystem projects and $210 billion in private investment have been catalysed by the CHIPS Act. Micron announced up to $200 billion in US manufacturing investment. Texas Instruments received $1.6 billion for three new facilities. Samsung committed $37 billion in Texas — the largest foreign direct investment in Texas history. TSMC's Arizona fabs, after well-publicised delays, are producing chips.
The Inflation Reduction Act added $369 billion for clean energy and climate manufacturing — EV batteries, solar panels, wind turbines — through a system of tax credits and manufacturing incentives. The combination has worked spectacularly for investment announcement purposes. Electrical equipment (EV batteries and solar) accounted for 31% of all reshoring jobs in 2024. Semiconductors took 35%. Two sectors alone drove two-thirds of the entire reshoring wave.
The honest challenge: TSMC's Arizona facilities encountered delays and cost overruns, pushing production timelines back. Building a cutting-edge fab from scratch in the US costs 40–50% more than in Taiwan. The CHIPS Act is succeeding at attracting investment announcements. The test is whether those announcements translate into operational fabs producing competitive chips at scale — and on what timeline. The bottleneck is not capital. It is the workforce to run the facilities. The US projected shortfall of 2.1 million manufacturing workers by 2030 is the binding constraint on the entire reindustrialisation agenda.
The EU Chips Act targets doubling Europe's share of global semiconductor production from 10% to 20% by 2030 — an extraordinarily ambitious target given that Europe's chip manufacturing base has been shrinking, not growing, for two decades. The Act has catalysed €69 billion across R&D projects and facility investments. TSMC announced a Dresden fab (€10 billion, German government co-funding). Intel began an ambitious Germany and Ireland expansion programme.
The Critical Raw Materials Act, upgraded through the RESourceEU action plan in December 2025, sets binding targets: by 2030, the EU must extract at least 10% of its annual consumption of strategic raw materials domestically, process at least 40% domestically, and source no more than 65% of any strategic material from a single third country. The 65% cap is aimed explicitly at China, which dominates the supply of rare earths, lithium processing, and critical semiconductor materials. Adopted by Council March 4, 2026, the regulation now requires the Commission to notify member states and company boards of any CRM supply risks — creating a formal early warning system for the critical minerals Europe's industrial future depends on.
The honest challenge: Europe faces a fundamental competitiveness problem that industrial policy cannot easily solve. Energy costs post-Ukraine are structurally higher than in the US or Asia. Labour costs are higher than in Asia. Regulatory complexity is higher than in the US. BASF moved production to the US precisely because the economics there work better. The EU can build the factories. The question is whether the products made in those factories can compete globally at a price the market will pay.
India's Production Linked Incentive (PLI) scheme — offering cash incentives of 4–20% of incremental revenue to manufacturers who meet production targets in India — is the most comprehensive attempt to accelerate domestic manufacturing since India's post-independence industrialisation. Running across 14 sectors including smartphones, pharmaceuticals, medical devices, textiles, food processing, solar PV modules, and white goods, the PLI has attracted commitments from Apple's supply chain (Foxconn, Pegatron, Wistron), Samsung, and dozens of domestic manufacturers.
The semiconductor mission is India's most ambitious industrial bet. Micron's $2.75 billion facility in Gujarat began partial operations in early 2025. Intel's glass substrate facility in Odisha is scheduled to commence operations in 2026. The Bharat Semi compound semiconductor fab is planned for 2027. India's goal: capture a meaningful share of the $1 trillion global semiconductor market by 2030. The US-India TRUST initiative establishes government-to-government coordination — positioning India as the democratic world's alternative to Taiwan for semiconductor assembly and advanced packaging.
The honest challenge: India's PLI has generated impressive investment announcements but uneven results in actual production ramp-up. The semiconductor ambition is genuine but the scale — $5–6 billion across the first wave of projects — is modest relative to the $20+ billion minimum for a single cutting-edge fab. India's path to semiconductor significance runs through assembly and packaging first, and advanced fabrication only later — a decade-long journey, not a five-year one. The infrastructure gap — power reliability, water supply, logistics — remains India's binding constraint on manufacturing scale.
The reindustrialisation is not uniquely American or European — it is a democratic-world industrial policy operating in parallel. Japan has committed over $150 billion in semiconductor subsidies, luring TSMC to Kumamoto (already producing chips ahead of schedule in 2024) and Samsung to a second Japanese facility. South Korea's K-Chips Act offers up to 25% tax credits for semiconductor facility investment — more generous even than the US CHIPS Act. Canada's Critical Minerals Strategy identifies 31 minerals vital to clean energy and defence and establishes a framework for allied supply chain coordination.
The most significant development is the emergence of coordinated allied industrial policy — the US, Japan, South Korea, the EU, Canada, and Australia explicitly coordinating to build supply chains that do not run through China for critical technologies. This is not protectionism against all trade — it is the construction of a trusted-partner supply chain architecture for the most strategic sectors: semiconductors, critical minerals, clean energy components, and defence equipment. The Chip 4 Alliance (US, Japan, South Korea, Taiwan) is the semiconductor expression of this; the Minerals Security Partnership is the critical minerals expression.
The 2025 USA Reshoring Survey asked 500 US manufacturers what would bring back the most manufacturing from offshore. The answer was not tariffs. Not a weaker dollar. Not lower corporate taxes. Not less regulation. A stronger skilled workforce would bring back more manufacturing than any of these — not marginally, but significantly. OEMs said they would reshore 30% of their products currently manufactured offshore if the skilled labour existed. Tariffs at 15% only brought back 23%. A 15% dollar depreciation: 21%. Corporate tax cuts to 15%: 18%.
The Deloitte and Manufacturing Institute projection is stark: 2.1 million manufacturing jobs will go unfilled by 2030 in the US alone — with a projected GDP impact of $1 trillion. This projection was made before the current wave of reshoring announcements added another quarter-million positions per year to the demand side. 88% of reshoring jobs in 2024 were classified as high-tech or medium-high-tech. Controls and automation engineers are the hardest roles to fill — consistently taking 60+ days to place even with active recruitment in 2025 and 2026. Manufacturing engineers for greenfield projects, where you're designing processes from scratch, are a close second.
This is not a problem that tariffs, subsidies, or policy declarations can solve quickly. Training a controls engineer takes four years of university followed by years of practical experience. Building the university programmes that produce them takes a decade of investment in curriculum, faculty, and laboratory infrastructure. The workforce problem has a structural time mismatch with the political urgency: politicians want factories producing in their term. Factories need engineers. Engineers need training that outlasts any political cycle. The countries that started investing in manufacturing workforce education ten years ago — Germany's dual apprenticeship system, South Korea's technical universities — have an advantage that cannot be replicated quickly.
Reindustrialisation and free trade are in fundamental tension — and the world hasn't yet decided which it wants more. The CHIPS Act, the IRA, and the EU Chips Act are all industrial subsidies that distort market outcomes — they make production in the subsidising country artificially cheaper than the market would otherwise choose. This is industrial policy, not free trade. The WTO framework that has governed global commerce since 1995 is being undermined by every major economy simultaneously. The question is not whether this tension can be avoided — it cannot. The question is whether the world can manage a transition from free-trade globalisation to strategic-industrial-policy globalisation without triggering a subsidy war that makes everyone worse off. The signs are mixed.
The cost of reindustrialisation is real and it will be paid by consumers. Manufacturing in the US or EU costs more than in Asia. That cost difference is partly absorbed by subsidies (paid by taxpayers) and partly passed through to consumers (paid by buyers of the products). A semiconductor made in Arizona costs 40–50% more to produce than one made in Taiwan. That premium is currently obscured by subsidies. When subsidies end or are reduced — as political cycles inevitably change — the premium emerges. The strategic case for reindustrialisation is genuine. But genuine strategic necessity does not make the economics painless.
The reindustrialisation wave connects directly to the ASEAN story in Letter 96. Every factory not built in China because of geopolitical risk is either built at home (reindustrialisation) or built in a trusted third country (ASEAN). These are not competing strategies — they are complementary ones. The US builds its most sensitive semiconductor fabs in Arizona. It builds its cost-competitive electronics assembly in Vietnam. Europe builds its critical minerals processing in Morocco. India builds its pharma and electronics at home. The reindustrialisation wave and the China+1 ASEAN wave are two expressions of the same underlying strategic reality: the world is reorganising its supply chains around geopolitical trust, not just economic efficiency. The investment opportunities are in every node of that reorganisation — not just the most visible ones.
Long-horizon thinking on capital, technology, and the forces shaping the next decade of wealth creation. Written from first principles. Not consensus. Not noise.