Malaysia's Manufacturing Sector: Bigger Every Year, Stuck at the Same Share of GDP

At a glance

Metric	Value	Period	Source
Share of GDP	~23.0%	2025	DOSM
Sector growth	4.5%	2025	DOSM
Whole-economy growth	5.2%	2025	DOSM
Jobs (registered private establishments)	2.54 million (27.5%)	Q4 2025	DOSM
Share of all vacancies	58.4% (115,800)	Q4 2025	DOSM
Labour productivity	RM59.0 per hour worked	2025	DOSM
Approved investment	RM131.3 billion (30.8% of total)	2025	MIDA
Peak GDP share	just under 31%	1999	World Bank

A quarter-century-old ceiling

Malaysia’s factories have never produced more output than they do now. Manufacturing expanded 4.5% in 2025, faster than the 4.2% recorded in 2024, inside an economy worth RM2.03 trillion at current prices.

And yet the sector’s share of GDP is almost exactly where it was in 2010.

That is the puzzle worth explaining, because most coverage of Malaysian manufacturing picks one half of it. Boosters cite record chip investment and call it a transformation. Critics cite the flat GDP share and call it stagnation. Both are reading the same number and neither is reading it carefully.

What the share actually did

World Bank national accounts data for Malaysia show manufacturing value added as a percentage of GDP moving like this:

Year	Manufacturing value added (% of GDP)
1995	26.4%
1999	30.9% (series peak)
2004	30.4%
2010	23.4%
2015	22.3%
2020	22.2%
2024	22.5%

The fall happened between roughly 2005 and 2010. Since then the line is essentially horizontal — the range across fifteen years is a little over one percentage point.

DOSM’s own 2025 figures put the sector in the same place. Services contributed 59.5% of GDP in 2025, and DOSM reports that services and manufacturing together contributed 82.5% — which leaves manufacturing at about 23.0%.

So this is a plateau, not a decline. Manufacturing output is growing; it is simply growing at close to the same rate as everything else.

Why a flat share is arithmetic, not necessarily failure

Three mechanisms produce a flat share, and they mean very different things.

Services grew faster. Services expanded 5.4% in 2025 against manufacturing’s 4.5%. When the larger sector grows faster, the smaller sector’s share falls even while its output rises. Every economy that gets richer sees this; the label for it is structural change, not deindustrialisation.

Value added is captured elsewhere in the chain. Malaysia’s position is concentrated in the back end of electronics — assembly, testing and packaging — where the margin per unit is thinner than in design or wafer fabrication. High export volumes therefore convert into less domestic value added than the headline trade numbers suggest. This is covered in detail in electronics and semiconductors and external trade.

Composition changed inside the sector. This is the part the GDP share cannot show you at all.

The evidence that the mix is shifting

If manufacturing were simply stuck doing the same low-margin assembly work at larger scale, productivity would be flat. It is not.

Productivity measure	Manufacturing	Services	All sectors
Value added per hour worked, 2025	RM59.0	RM42.0	RM44.5
Growth, 2025	+4.6%	+3.3%	+3.7%
Value added per employed person, 2025	RM139,454	—	RM102,672

Manufacturing generates about 40% more value per hour than services and about a third more than the national average. That is not the profile of a pure assembly economy.

Sub-sector growth in 2025 points the same way. Electronic components, communication equipment and consumer electronics grew 9.9% for the year, and in the fourth quarter alone electrical, electronic and optical products surged 12.7%. Food processing grew 9.1%, and vegetable and animal oils and fats 9.9% — the commodity-linked side of manufacturing that connects to the palm oil sector.

Investment approvals tell a similar story. MIDA approved RM131.3 billion of manufacturing investment in 2025 — 30.8% of Malaysia’s record RM426.7 billion total — across 1,354 projects expected to create 109,950 jobs. The largest industries by approved value were E&E (RM28.5 billion), chemicals and chemical products (RM24.9 billion), transport equipment (RM14.9 billion), basic metal products (RM11.1 billion) and machinery and equipment (RM11.0 billion). MIDA’s Managerial, Technical and Supervisory index — the skilled share of promised positions — rose to 42.8%.

Note also the ownership split: RM100.6 billion (76.6%) of that manufacturing investment was foreign, RM30.7 billion (23.4%) domestic. Upgrading is happening largely on foreign balance sheets.

Jobs: the number people misquote

DOSM’s Quarterly Employment Survey counted 2.54 million manufacturing jobs in Q4 2025, or 27.5% of all jobs it tracks, with 2.42 million of them filled. Manufacturing also accounted for 58.4% of every unfilled vacancy in the survey — 115,800 positions.

That vacancy concentration is the single most revealing labour statistic in the sector. It says employers cannot fill roles at the wages and conditions on offer, which is exactly what you would expect from an industry mid-transition: the old jobs are hard to staff and the new ones are hard to staff for different reasons.

Read the 27.5% carefully, though. See the common mistakes below.

The middle-income trap argument, stated fairly

The trap thesis holds that a country can grow rich enough to lose its low-cost advantage without becoming innovative enough to compete on technology, and then sits between the two. Malaysia is the case study most often cited.

The strongest evidence for it is not the GDP share. It is the wage share. Compensation of employees was 33.6% of GDP in 2024 (RM648.5 billion), up only marginally from 33.5% in 2023, while gross operating surplus took 63.7%. In high-income economies the wage share typically sits far higher. Output can rise for years without much of it reaching households.

The government’s own planning documents accept the framing. The Thirteenth Malaysia Plan (2026-2030) targets:

• GDP growth of 4.5% to 5.5% per year
• Manufacturing growth of 5.8% per year — deliberately above services at 5.2%
• Gross exports growth of 5.8% per year
• GNI per capita of RM77,200 by 2030, which the Plan states would surpass the high-income threshold
• Compensation of employees rising to 40% of GDP by 2030
• Foreign workers reduced to an average of 10% of the workforce by 2030, and up to 5% by 2035

DOSM reports GNI per capita reached RM57,200 in 2025, up from RM54,870 in 2024. The gap to RM77,200 in five years is the whole policy problem in one line.

The instruments named in the Plan are the New Industrial Master Plan 2030, the National Semiconductor Strategy and the National Energy Transition Roadmap. Note what the manufacturing target implies: for manufacturing to grow at 5.8% while the economy grows at 4.5-5.5%, its share of GDP has to start rising again. The plateau ending is not a forecast — it is a target.

Common mistakes

Saying manufacturing employs 27.5% of Malaysians. It does not. The Quarterly Employment Survey covers registered private-sector establishments — 9.21 million jobs in Q4 2025. It excludes the public sector, the self-employed and informal work. Manufacturing’s share of everyone in work is materially lower. Use the Labour Force Survey for that question; see unemployment and the labour force.

Treating the falling GDP share as deindustrialisation. The fall ended around 2010. Confusing a one-off structural adjustment with an ongoing decline misdates the problem by fifteen years.

Using export value as a proxy for manufacturing value added. Electronics exports can rise far faster than the domestic value added behind them, because much of the imported input value passes straight through. Check the value-added series in the national accounts rather than inferring it from trade headlines.

Quoting approved investment as realised investment. MIDA’s RM131.3 billion is approvals. Realisation lags by years and not every approved project proceeds.

Comparing the 23.0% share to China or Germany. Different economies book different activities inside “manufacturing”; contract-manufacturing hubs and brand-owning economies are not measuring the same thing.

What’s next

Watch four series rather than the headline share. DOSM’s quarterly labour productivity release shows whether the value-per-hour gap over services keeps widening. The annual GDP income-approach release shows whether the wage share moves off 33-34%. MIDA’s quarterly approvals show whether the domestic share of manufacturing investment rises above a quarter. And the sub-sector detail in each quarterly GDP release shows whether growth is broadening beyond electronics.

For the wider picture, start with the GDP overview, then read electronics and semiconductors for the sector that dominates both manufacturing output and external trade.