17.5 against 40.7
Start with the ratio, because it is the one that does not get quoted.
By: Robert Sterling – AsiaPresswire – In the first quarter of 2025, non-oil and gas manufacturing accounted for 17.50% of Indonesian GDP — Rp 991,785.2 billion of output at current prices, according to Statistics Indonesia. In the twelve months to the second quarter of 2025, the same sector absorbed 40.72% of total realised national investment: Rp 568.4 trillion, according to Industry Minister Agus Gumiwang Kartasasmita’s October 2025 press conference.
Divide one by the other. Manufacturing takes about 2.33 times its share of GDP in investment.
That is the whole argument, in one ratio. A sector producing roughly a sixth of output is consuming roughly two fifths of the capital being deployed in the country. Either that is a deliberate bet on industrialisation that will pay off later — which is the government’s position, and a defensible one — or it is capital being spent at a poor rate of conversion. Both readings are compatible with the data available. The data available does not, on its own, settle which is true.
The third number in the set is jobs. That same Rp 568.4 trillion of manufacturing investment is credited with 19.44 million workers absorbed, 13.41% of national employment. So: 40.7% of the capital, 17.5% of the output, 13.4% of the jobs.
Read in that order, the sector is capital-hungry relative to what it currently produces and to whom it currently employs. That is what a country in the middle of a downstreaming push looks like. It is also what a country with an expensive logistics base looks like. The rest of this piece is about telling those two apart.
Eight months of PMI, read in order
The monthly sentiment data is where the volatility lives, and reading it as a series rather than as a headline is the entire trick.
S&P Global’s Indonesia Manufacturing PMI, 2025:
- January: 51.9 — moderate expansion, the highest reading since May 2024.
- February: 53.6 — strongest growth in eleven months, driven by new orders and output.
- March: 52.4 — slower than February, still comfortably above 50.
- April: 46.7 — sharp contraction, the steepest since August 2021.
- May: 47.4 — still contracting, slightly less badly.
- August: 51.5 — back above the line.
- September: 50.4 — slower growth.
- October: 51.2 — third consecutive monthly improvement, with employment rising at the steepest pace in five months.
The range across eight months is 6.9 points. That is not a sector gently trending. That is a sector whose monthly readings swing between “strongest in nearly a year” and “worst in nearly four years” inside ten weeks.
Two things sit behind the April collapse. Domestic demand weakened after the post-Eid consumption period, and export orders to the United States fell as tariff uncertainty hit. Layoffs in the wider economy passed 24,000 workers by May 2025, which feeds straight back into domestic demand for manufactured goods. Input costs rose at the same time, and manufacturers absorbed most of it: factory gate prices rose only fractionally, because raising them would have cost volume.
October’s reading is the more informative one, because of what did not happen. New orders rose for a third month at a pace equal to the joint-strongest since March. Employment rose. But production was flat, and panel members attributed that partly to having run down finished-goods inventory in prior months. Export orders fell for the second consecutive month.
Usamah Bhatti, economist at S&P Global Market Intelligence, put it this way in the October release: “The improvement in the health of the Indonesian manufacturing sector strengthened at the start of the fourth quarter of 2025, boding well for the coming months.” He also flagged the price side — “Firms were hesitant to fully pass through higher expenses to clients, as factory gate charges rose only fractionally as part of efforts to remain competitive on price.”
The quarter-by-quarter share series
The GDP share figures are more stable than the PMI and more useful than any single annual number, because they show the direction.
Non-oil and gas manufacturing as a share of GDP at current prices, by quarter:
- Q1 2024: 17.47% (Rp 923,722.5 bn)
- Q2 2024: 16.70% (Rp 924,737.8 bn)
- Q3 2024: 17.18% (Rp 968,780.8 bn)
- Q4 2024: 17.31% (Rp 982,048.2 bn)
- Q1 2025: 17.50% (Rp 991,785.2 bn)
The spread is 0.8 percentage points across five quarters. The sector is not collapsing and it is not surging; it is holding a share of roughly 17% to 17.5% while the rest of the economy grows around it.
There is a longer arc behind those five quarters, and it points the other way. Annual estimates put manufacturing’s share near 19.3% in 2020, then 18.9%, 18.6%, 18.7%, around 18.5% in 2024, and a projected 18.2% for 2025. On that measure, Indonesian manufacturing has been losing share for five years.
The two series do not contradict each other. The quarterly BPS figures measure non-oil and gas manufacturing; the annual series is a broader definition. But the direction they jointly suggest is that Indonesia is holding its factory base roughly flat as a share of the economy, against a trend of slow erosion that predates the current administration.
Growth rates sit alongside this. The minister put sector growth at 4.94% from the fourth quarter of 2024 to the second quarter of 2025. BPS put non-oil and gas manufacturing growth at 4.31% year on year in Q1 2025, with food and beverage expanding 6.04% — the fastest large sub-sector, and the one that contributes over 30% of manufacturing GDP.
Splitting Rp568.4 trillion by worker
Per-unit arithmetic is where big national figures stop being impressive and start being checkable.
Take Rp 568.4 trillion of annual manufacturing investment against 19.44 million workers in the sector. That is roughly Rp 29.2 million per employed person, or about USD 1,780 at an exchange rate in the region of 16,000 rupiah to the dollar.
That figure needs two caveats before anyone uses it. The Rp 568.4 trillion is a flow — one year of realised investment. The 19.44 million is a stock — everyone currently employed in the sector. Dividing a flow by a stock gives a ratio, not a cost per job created, and the real cost per *new* job is several times higher. The rupiah figure is also sensitive to the exchange rate, and 2025 was not a stable year for it.
With those caveats, the ratio still tells you something: Indonesia is deploying capital into manufacturing at a rate of roughly Rp 29 million per person already working there. Compare that with a sector where the median manufacturing wage is a small fraction of that figure, and the gap is the capital intensity of downstreaming. Smelters, refining and chemical processing are not labour-intensive. They are capital-intensive by design, and Indonesia chose them deliberately.
Per-worker output tells the same story from the other side. Quarterly manufacturing output of roughly Rp 991,785 billion across 19.44 million workers is about Rp 51 million per worker per quarter, or roughly Rp 204 million a year — around USD 12,700 of gross output per person. Again: gross output, not value added, and not a wage. But it is the number to hold when someone tells you Indonesian manufacturing is low-productivity, and it is the number that makes the logistics figure below hurt.
25% exported, 75% sold at home
The minister made a point in October 2025 that deserves more attention than the growth number: Indonesian manufactured goods go 25% to export markets and 75% to the domestic market.
This is the structural fact that separates Indonesia from its neighbours. Vietnam and Thailand run export-led manufacturing sectors. Indonesia runs a domestically-absorbed one. The trade data looks weak by comparison — the minister acknowledged Indonesian exports trail Malaysia, Vietnam and Thailand — and the explanation offered was that this is a market-orientation choice rather than a competitiveness failure.
Both things can be true. A large domestic market is a genuine advantage: 280 million people absorb a great deal of what the country makes, and that demand does not vanish when a foreign tariff changes. It also means the sector is exposed to domestic purchasing power in a way Vietnam’s is not, which is precisely what the April 2025 PMI collapse demonstrated. When domestic demand weakened, factories cut output within weeks.
The export share is where the policy energy is going. Hilirisasi — downstream processing — is explicitly designed to move Indonesia up the export value chain: not nickel ore but nickel products, not bauxite but aluminium. Central Sulawesi and North Maluku have become investment hotspots on exactly this logic.
- Food and beverage: over 30% of manufacturing GDP, growing 6.04% in Q1 2025, overwhelmingly domestic.
- Automotive: driven by domestic demand and EV-related investment, concentrated in West Java.
- Chemicals and pharmaceuticals: expanding on health demand and industrial inputs.
- Textiles and apparel: facing export competition, pivoting toward sustainable fibres.
- Electronics: still modest, growing on regional supply chain shifts.
Only two of those five are meaningfully export-oriented today.
23.5% of GDP spent on moving things
Here is the number that explains the most and gets discussed the least.
Indonesia’s logistics cost is estimated at 23.5% of GDP, against 16.8% for Vietnam and 14.5% for Thailand, on World Bank figures. That is a gap of roughly 6.7 percentage points against Vietnam and 9 points against Thailand.
Put it against the sector it bites hardest. If manufacturing is about 17.5% of GDP, then a logistics burden nine points above a direct competitor is a tax on every factory that has to move anything — which is all of them. On a per-unit basis it is the difference between a West Java plant being competitive for an export order and not being competitive, before a single wage or tariff is considered.
Why is it so high? The answers are structural and unglamorous:
- An archipelago with a road network concentrated on Java and inter-island movement dependent on shipping.
- Port dwell times and handling costs that are improving but remain above regional comparators.
- A trucking sector with low average fleet age and high empty-running rates.
- Warehousing concentrated around Jakarta, so anything made elsewhere travels to Java before it travels anywhere else.
That last point interacts with the next section. Manufacturing is concentrated in Java partly for good reasons — infrastructure, market proximity, suppliers — and that concentration then generates logistics costs for everyone outside Java who wants to participate in the same supply chains.
170 parks, 95,000 hectares, 59.5% full
As of May 2025, 170 industrial park operators held licences, covering close to 95,000 hectares, with an occupancy rate of 59.52%.
Sixty per cent occupancy is the number to sit with. It means roughly 38,000 hectares of serviced, licensed, industrial-zoned land is currently empty. That is not necessarily a bad thing — a park needs vacancy to absorb a large new tenant, and 40% vacancy is not unusual for a young estate — but it does complicate the argument that Indonesia’s constraint is a shortage of industrial land.
If 40% of licensed park land is empty, the binding constraint is somewhere else: power, water, port access, or the cost of getting a tenant through licensing fast enough to matter. The Ministry of Industry’s response has been to draft a regulation for Special Industrial Zones, or KIT, intended for smaller or themed parks in land-constrained or strategically located areas. That is a sensible instrument for the land-constrained case. It does nothing for the empty-38,000-hectares case.
The occupancy figure also hides enormous variation. A park in Bekasi or Karawang near Jakarta will be near full. A park in Central Sulawesi attached to a nickel project will be full by definition. A park in a secondary province without a dedicated power plant or a port agreement will sit at a fraction of that, and will sit there for years.
Six months in Java, eighteen in Kalimantan
The regional distribution is the most concentrated part of the whole picture.
Java hosts more than 75% of Indonesia’s medium- and large-scale manufacturing establishments. West Java alone accounts for over a third of the national total, by virtue of sitting next to Jakarta with access to ports and a dense supplier base. Sumatra and Sulawesi are growing, mostly on resource processing. Kalimantan is growing from a very low base.
Then there is the timeline problem. Land acquisition for industrial use takes roughly six months in Java. In Kalimantan or Papua, it can run beyond eighteen months, driven by conflicting local land-use rules and overlapping claims.
An eighteen-month differential is not an inconvenience. It is the difference between catching an investment cycle and missing it. A company evaluating two sites will discount the second one by the carry cost of eighteen months of capital, plus the risk that the land is still not clear at the end of it. That is how a country ends up with 75% of its factories on one island while officially pursuing regional equity.
Layer on the licensing picture. The Omnibus Law was written to simplify exactly this, but reporting cited by industry analysts indicated fewer than 40% of provinces had fully implemented its provisions by 2023, per Ministry of Law and Human Rights data. A national reform that lands in under half the provinces produces a country where the answer to “how long will this take” depends entirely on which map you are standing on.
Where the capital came from
Manufacturing foreign direct investment in the first quarter of 2025 totalled USD 28.2 billion, up about 14% year on year.
The source list is a map of the region’s capital flows: Singapore, China, Hong Kong, South Korea and Japan. Destination hotspots by commodity: Central Sulawesi for nickel, North Maluku for smelters, West Java for automotive and electronics.
Twenty-eight point two billion dollars in a quarter is a large number, and it is worth noting what it mostly is: capital for processing facilities, not for light manufacturing employment. A nickel smelter in Central Sulawesi might employ a few thousand people for a billion dollars of investment. A garment factory in West Java might employ the same number for a few tens of millions.
That is not a criticism of downstreaming. It is an explanation of the 40.7%-investment / 13.4%-employment pair at the top of this piece. Indonesia is choosing capital-intensive processing because it has the ore and it wants the value added, and it is accepting a lower employment yield per dollar as a result.
Two policy levers are currently in play. The Domestic Component Level, or TKDN, is being relaxed, with President Prabowo Subianto asking his administration to ease requirements to attract more foreign investment. And the KIT regulation is being drafted to allow smaller, themed parks. Both push in the same direction: make it easier for foreign capital to land. Neither addresses the logistics cost.
What 6.9 points of swing costs a factory
The PMI is a diffusion index, not a volume index, and it is worth translating what a 6.9-point range actually does to a factory floor.
A reading of 53.6 in February and 46.7 in April is not “slightly better” and “slightly worse.” Above 50, more firms report improvement than deterioration; below it, the reverse. April’s 46.7 was reported as the steepest contraction since August 2021, which means a majority of panel firms reported conditions deteriorating, and deteriorating at a rate not seen in nearly four years.
For a factory manager, that translates into three concrete costs.
- A sector that swings this fast has to hold more buffer stock than one that trends. Buffer stock is working capital, and working capital in Indonesia is not cheap.
- Output flat while employment rises — the October pattern — means labour is being held through a soft patch rather than shed. That is good for the workers and expensive for the balance sheet, and it is visible in the October data as the steepest employment rise in five months alongside unchanged production.
- Input costs rose at the strongest rate in eight months in October while factory gate prices rose only fractionally. Margin compression of that kind does not show up in GDP. It shows up in the next investment decision.
The range also matters for anyone modelling the sector. A single point estimate for Indonesian manufacturing growth in 2025 is close to meaningless when the monthly sentiment data spans 46.7 to 53.6 inside one year. Any credible forecast has to be a range, and the range has to be wide — roughly four to five percent growth if domestic demand holds, and something nearer flat if the post-Eid pattern of 2025 repeats and export orders to the United States stay soft.
That is the honest position. The sector is expanding, the employment signal turned positive in October, and the swing is large enough that anyone claiming precision about the next two quarters is guessing.
Two numbers to check next quarter
The argument about Indonesian manufacturing will not be settled by the PMI, which will keep swinging, or by the GDP share, which moves in fractions of a point.
Check these two instead.
Occupancy rate of licensed industrial parks. It was 59.52% in May 2025. If it rises, the capital is landing and the downstreaming bet is converting into operating plants. If it stays flat while investment keeps rising, the money is going into announcements and land banks rather than production. This is the single cleanest indicator available and it is published.
Manufacturing share of GDP at current prices. It was 17.50% in Q1 2025, up from 16.70% a year earlier. If the next four quarters hold above 17.5%, the sector is genuinely growing faster than the economy. If it drifts back toward 17% and then below, the 40.7% investment share is buying market share in nothing.
Everything else — the logistics percentage, the land acquisition timelines, the TKDN negotiation — is the mechanism. These two are the readout.
source https://newsroom.seaprwire.com/press-releases/finance/three-numbers-that-decide-indonesias-factory-story/
