A choice most rich economies did not make
Most advanced economies let manufacturing shrink to a rounding error. Singapore did not. In 2025 the sector produced 18.5% of gross domestic product — a share the United States, Britain and Australia gave up decades ago — and it grew 8.7% in real terms while doing so. In the second quarter of 2026 manufacturing expanded 12.2% year on year, the single largest contributor to the economy’s 5.7% growth. The question is not whether Singapore still makes things. It is how a city-state with no natural resources, six million people and some of Asia’s highest land and labour costs remains a factory floor for the world.
Why Singapore still makes things at scale
The short answer is that it stopped competing on cost around 1990 and started competing on capability. Singapore does not win the assembly of cheap goods; it wins the fabrication of hard ones — wafers, active pharmaceutical ingredients, jet-engine components, offshore rigs. These are products where a defect is expensive, intellectual property is the asset, and the buyer pays for reliability rather than the lowest quotation. That positioning is deliberate policy, executed over three decades by the Economic Development Board (EDB), and it is why the 18.5% figure has proved durable rather than a statistical accident.
The scale is real. Singapore supplies about one in ten of the world’s semiconductor chips and roughly a fifth of the world’s semiconductor manufacturing equipment. It is a significant global exporter of pharmaceuticals by value, and a leading builder of jack-up rigs and floating production vessels. For a country smaller than New York City, the concentration of high-value output is unusual.
The 18.5% story, and how it held
Across the OECD, manufacturing’s share of output has trended toward 10% or below. Singapore’s has held near a fifth. It managed this by treating the sector as a portfolio to be actively managed, not a legacy to be defended. When textiles and simple electronics assembly became uncompetitive in the 1980s and 1990s, the state did not subsidise them — it courted the next tier of activity and let the old tier migrate to Malaysia, Indonesia and beyond. The result is a base weighted toward the value-dense end of every industry it retains.
The 2025 numbers show the model working. Manufacturing real value-added grew 8.7% for the full year. In the second quarter of 2026 it accelerated to 12.2% year on year — outpacing construction and the trade-related services cluster. When manufacturing leads the economy rather than dragging on it, the case for keeping factories is easy to make.
The six clusters and what each contains
Singapore’s official statistics divide manufacturing into six clusters, and the full-year 2025 growth rates reveal a sector moving at very different speeds.
Electronics is the anchor. It grew 12.7% in 2025 and accounts for 43.2% of all manufacturing value-added — equivalent to 8.0% of Singapore’s entire GDP. This cluster is semiconductors first: wafer fabrication, chip design, advanced packaging, and the specialised equipment that makes chips.
Transport engineering grew 18.7% in 2025, the fastest of the six. It spans aerospace — Singapore is Asia’s leading maintenance, repair and overhaul hub — and marine and offshore engineering, where the rig and vessel order book has recovered.
Biomedical manufacturing grew 16.3% in 2025. It covers pharmaceuticals, biologics and medical technology; output is famously lumpy because a single high-value batch can swing the monthly index.
Precision engineering grew a steadier 3.6%. This is the supplier spine — precision machining, tooling, moulds, automation, analytical instruments — that feeds the other clusters and the region’s factories.
Chemicals was flat, up 0.2%, reflecting a mature refining and petrochemical base on Jurong Island facing thin margins and energy-transition pressure.
General manufacturing — food, printing and miscellaneous products — contracted 7.8%, the weakest of the six and a reminder that the low-value tail continues to erode.
A two-speed base
Read together, the cluster figures describe a two-speed sector. One speed is semiconductor-led electronics and precision engineering, riding structural demand for AI compute, memory and the equipment that produces both. The other is the cyclically or structurally exposed set — chemicals grinding against margins, general manufacturing in secular decline, biomedical swinging batch to batch.
This matters for reading headline growth. The 12.2% expansion in the second quarter of 2026 was disproportionately an electronics story; chemicals and biomedical output declined in the same quarter. Singapore’s manufacturing strength is genuine but concentrated. A downturn in the global chip cycle would be felt here more sharply than the aggregate number suggests, which is why the state keeps trying to broaden the base rather than lean harder on one cluster.
Moving up the value curve, not out
The strategic through-line is consistent across clusters: go up, not out. In semiconductors that means advanced-node logic, high-bandwidth memory and advanced packaging rather than commodity assembly. In pharmaceuticals it means biologics and complex active ingredients rather than generic tablets. In precision engineering it means five-axis machining and metrology rather than simple stamping. Singapore is more expensive than every regional alternative; it competes on capability, not price, and capability is the only ground on which it chooses to fight.
Employment and productivity
Manufacturing employed 494,400 people at December 2025, about 13% of the workforce excluding migrant domestic workers. That is a smaller headcount than the sector’s output share implies — the point of moving up the value curve is that each worker produces more. Singapore’s manufacturing output per worker is among the highest in the world, a function of automation, capital intensity and the deliberate shedding of labour-heavy activity.
The trade-off is visible in investment forecasts: the jobs promised per dollar of investment have fallen as projects grow more automated. EDB’s 2025 commitments are expected to create 15,700 jobs over five years, and a large share are expected to pay well above the median wage. Singapore is buying productivity and wages, not headcount.
The EDB investment engine
The Economic Development Board is why the pipeline stays full. In 2025 it secured S$14.2 billion in fixed-asset investment commitments, up from S$13.5 billion in 2024, plus S$8.9 billion in total business expenditure. Manufacturing-related projects accounted for the bulk of the fixed-asset total, with electronics and biomedical drawing the largest shares — the two clusters the state most wants to deepen.
A striking shift sits inside the 2025 numbers: the source of investment is diversifying. The United States’ share of fixed-asset commitments fell sharply while China’s rose, and Chinese firms accounted for roughly half of business expenditure — evidence of a two-way de-risking that the companion China+1 article explores. The source changed; the total held.
The supplier spine and the agencies behind it
Beneath the marquee multinationals sits a dense layer of small and mid-sized suppliers — precision machinists, tool-and-die shops, surface-treatment specialists, automation integrators — that make the ecosystem function. This spine is why an anchor plant can localise a supply chain rather than fly parts in, and it is the most fragile part of the base.
Four agencies steer the system. EDB attracts and anchors the multinationals. Enterprise Singapore builds the local suppliers and pushes them overseas. JTC develops the industrial land and specialised facilities — Jurong Island for chemicals, wafer-fab parks for semiconductors. A*STAR supplies the public research that keeps capability ahead of lower-cost rivals. The division of labour is one of the least glamorous and most effective parts of the model.
Outlook
The near-term signal is positive but not euphoric. The SIPMM manufacturing PMI stood at 51.3 in June 2026 — above the 50 line that separates expansion from contraction, but a long way from a boom reading. The AI-driven semiconductor cycle is doing the heavy lifting; chemicals and general manufacturing remain soft.
The medium-term case rests on whether Singapore keeps winning capability that cannot be easily replicated elsewhere. On current evidence — the foundry commitments, the packaging plants, the biologics lines — it is. The risk is concentration: a sector 43.2% weighted to electronics is exposed to one global cycle. The opportunity is that the world’s push to de-risk supply chains is sending exactly the high-value work Singapore is built for. For a country that chose to keep making things, 2025 and 2026 have vindicated the choice.