# Two routes to industrial expansion

Comparative Study · Industrial Economics
Industrial expansion can begin by improving the connections among existing producers or by establishing another stage in a resource-processing chain. The first route seeks to reduce the cost and uncertainty of moving inputs and output. The second seeks to perform more transformation within the domestic economy. Both require capabilities beyond the visible asset, and neither can be assessed adequately from expenditure or export receipts alone.

India's coordinated infrastructure planning and Indonesia's resource-downstreaming strategy provide two distinct policy settings. They should not be treated as exhaustive descriptions of either economy: India also processes resources, and Indonesia also builds logistics infrastructure. The comparison concerns the mechanism under examination, rather than assigning each country one industrial model.

The common benchmark is the additional domestic production and income enabled by a project, after identifying imported inputs, operating requirements, capital use and public costs. This study develops separate, reproducible accounts for infrastructure and processing. The official evidence identifies actual program design and historical developments; the numerical project accounts are hypothetical and do not estimate the returns on either country's policies.

## India: coordination is different from completed construction

India's Ministry of Commerce and Industry describes PM GatiShakti's Network Planning Group as a mechanism for integrated infrastructure evaluation, including multimodality and last-mile connectivity. Its August 5, 2025 parliamentary reply reports that 293 projects, with a stated value of 13.59 lakh crore rupees, had been evaluated. It also states that the master plan has no separate specific fund allocation. These are planning and institutional observations, not a claim that the entire evaluated value had been spent or that all projects were operating. [PIB, PM GatiShakti project evaluation](https://www.pib.gov.in/PressReleasePage.aspx?PRID=2152521&lang=2&reg=48).

The distinction is economically important. A shared planning platform can identify missing links, overlapping works or a mismatch between a production site and transport capacity. Its potential contribution is to improve the coordination of investment. The number of projects evaluated does not independently establish the amount of travel time saved, the reduction in inventory costs or the resulting industrial output.

An operational evaluation should connect the planning decision to the completed network and then to users' outcomes. A port expansion without a reliable road or rail connection can leave a different constraint binding. A corridor with adequate physical capacity can still face scheduling or documentation delays. The mechanism being tested is the complete movement of goods, not only the existence of a new structure.

## Indonesia: processing changes the production boundary

OECD's 2024 Indonesia survey describes incentives and restrictions intended to encourage domestic processing, notably nickel. It records ferronickel and processed-nickel export values rising from 4.5 billion dollars in 2019 to 19.6 billion in 2022. The survey also calls for evaluation of costs, benefits and capacity risks. These are historical observations and the OECD's stated assessment; they do not establish the net social return of every processing project or the current legal status of each restriction. [OECD, Indonesia Economic Survey 2024, downstreaming discussion](https://www.oecd.org/content/dam/oecd/en/publications/reports/2024/11/oecd-economic-surveys-indonesia-2024_e3ab8960/de87555a-en.pdf).

The increase in export value establishes a substantial change in recorded processed shipments over that interval. It combines prices, quantities and the composition of products. A domestic-income assessment additionally needs the origin of feedstock, energy and other inputs, the contribution of resident production, depreciation and the distribution of operating income.

A processing plant also depends on transport and power. The two policy routes therefore overlap in their complementary inputs. Their distinct starting mechanisms do not eliminate the need for the same basic operating capabilities: reliable utilities, appropriate workers, maintenance, finance and access to customers.

## Compare the constraints being removed

| Dimension | Infrastructure coordination | Additional resource processing |
| --- | --- | --- |
| Immediate constraint examined | Cost, reliability or connectivity of moving goods | Missing or limited transformation stage |
| Main operating evidence | Journey time, variability, throughput and user costs | Yield, energy intensity, input sourcing and saleable output |
| Domestic capability | Network operation, logistics and coordination | Process engineering, maintenance and quality control |
| Principal demand question | Whether users employ the improved route | Whether customers buy the processed product at a viable margin |
| Main accounting caution | Construction cost is not the benefit delivered | Gross export value is not domestic value added |

The matrix establishes comparable questions without forcing identical project metrics. A road's output cannot be measured in metric tons of refined material, while a smelter's performance cannot be inferred from kilometers of connected routes. Both can be assessed through the resources they use and the domestic activities they enable.

## A transparent infrastructure account

Consider an invented freight network carrying one hundred shipments each year. Assume each shipment previously used logistics resources costing five monetary units. An improvement reduces that requirement to three for an otherwise identical service. The annual gross resource saving is two hundred.

Now assume that operating and maintaining the improvement uses fifty units annually and its consumption of fixed capital is another fifty. Under these restricted assumptions, the annual net resource saving is one hundred. The calculation is `100 × (5 − 3) − 50 − 50`. It excludes changes in shipment volumes, safety, emissions, land use and travel-time variability.

The expenditure constructing the improvement is not added again as an annual benefit. Nor is the gross saving necessarily a rise of the same amount in measured GDP. A cheaper logistics service can reduce expenditure on transport while supporting other productive activity. The welfare and output effects require a fuller production and use account.

The distribution is also unresolved. Lower operating costs may become lower freight charges, higher logistics-company margins or some combination. The result depends on competition, contracts and demand. A project-benefit estimate should identify who gains rather than assume that every saving reaches manufacturers in full.

For India, the relevance of the example is the evaluation sequence. Project coordination can be assessed against a specific missing connection and observable service improvement. The hypothetical savings do not validate any actual project. They identify the user and operating evidence needed to move from an evaluated investment amount to an economic result.

## A processing account with the same discipline

Take a separate hypothetical resource chain. Before an additional processing stage, an exported resource is worth one hundred, containing domestic value added of eighty and imported inputs of twenty. After processing, the export is worth one hundred sixty. The processor uses the original resource, domestic services invoiced at thirty, directly imported energy of twenty and its own value added of ten.

Assume the domestic services contain imported inputs of ten and resident value added of twenty. Total domestic value added in the processed export is then one hundred ten: eighty from the resource stage, twenty from domestic services and ten from processing itself. Foreign value is fifty: twenty already embodied in the resource, ten in the service inputs and twenty in direct energy imports.

| Hypothetical chain | Before processing | After processing |
| --- | ---: | ---: |
| Export value | 100 | 160 |
| Domestic value added | 80 | 110 |
| Foreign value added | 20 | 50 |

Export value rises by sixty, while domestic value added rises by thirty. If the additional domestic activities consume fixed capital worth fifteen each year, the increase in net domestic value added is fifteen before other adjustments. Neither increase is an estimate for Indonesia. The exhibit isolates the distinction between an additional production stage and the value retained within the domestic production boundary.

Public subsidies, environmental effects and alternative uses of the resource have not been included. A complete policy assessment would compare those costs and the counterfactual destination of domestic inputs. The larger export invoice is therefore evidence of more transformation under the example, not a complete benefit-cost result.

## Complementary inputs can determine the return

Both routes can fail to achieve their intended effect when a complementary input remains constrained. A logistics improvement may deliver little industrial gain if a factory cannot obtain reliable power. A processing plant may remain below capacity if transport cannot supply feedstock or move finished output. The relevant production system includes those connections.

Skills are similarly specific. Building a facility does not establish the ability to operate it safely, maintain it or meet the customer's specification. Importing specialist services can support initial operations, while developing resident capability can change the domestic contribution over time. These should be observed rather than assumed from the nationality of the investor.

The timing of complementary investments matters. Completing an asset before its required network connection can increase financing and maintenance costs during an idle period. Completing the connection after customers have committed to another supplier can reduce the available market. A project schedule should therefore align dependencies rather than report each completion date as an independent success.

Institutional coordination is not costless. Agencies and firms can face different budgets, incentives and information. The relevant test is whether coordination changes a specific decision and produces an observable service or production improvement. The presence of a committee or digital platform is evidence of an arrangement, not proof of its economic effect.

## Price and demand risks differ across the routes

A processing margin depends on the relationship between finished-product prices and feedstock, energy and other costs. A rise in the commodity price can increase export revenue while compressing a processor's margin if input costs rise more. Production volumes, prices and profitability should therefore remain separate observations.

Infrastructure has demand risk too. A route built for projected industrial traffic can be underused if the expected plants are delayed or demand shifts. The same construction cost can yield different unit costs depending on utilization. A forecast of traffic should state the facilities, production and customer assumptions that support it.

For either route, an optimistic demand projection can make a weak project appear attractive. A useful sensitivity exercise varies the specific demand or price condition that determines viability. It should preserve the operating and financing constraints instead of assuming full use under every scenario.

The public balance sheet adds another exposure. Guarantees or support payments can move risk from a project company to the government without changing the underlying resource requirement. Evaluation should identify that transfer and its conditions. Announced private investment is not independently evidence that public contingent risks are negligible.

## Capability and broader industrial effects

Infrastructure can support many users beyond the first major project. A processing facility can create demand for specialized maintenance and engineering. These possibilities are relevant but require evidence of actual use, supplier activity and resulting productivity. They should not be counted as fixed multipliers attached to every unit of expenditure.

An assessment of domestic capability should distinguish the value of current purchases from learning that improves future performance. A supplier can sell more of an unchanged service without developing a more advanced capability. Alternatively, a demanding contract can support a lasting improvement. The difference requires operating outcomes and a credible comparison, rather than an assumption that all local procurement is a spillover.

Distribution across regions and workers also matters. A project can generate national benefits while concentrating disruption, environmental burdens or adjustment costs locally. The production account should be supplemented with those effects when assessing policy. They cannot be resolved by the aggregate export or investment total.

## The comparative conclusion

The Indian evidence identifies a coordination mechanism and a documented project-evaluation pipeline. The Indonesian evidence identifies a historical shift toward resource processing and the need to assess its domestic contribution beyond export receipts. Those settings share a dependence on complementary inputs while addressing different immediate constraints.

The two worked accounts show how to keep those differences economically comparable. Infrastructure should be connected to service improvements and resource costs. Processing should be connected to direct and indirect inputs, domestic value added and capital consumption. Both then require a counterfactual and a record of public and external costs before a net policy return is claimed.

Industrial expansion is consequently more than constructing an asset or exporting a more processed product. The durable contribution depends on whether the investment removes a binding constraint, operates with the necessary capabilities and generates income that justifies the resources used. Applying that test to both routes supports a more informative comparison than treating either a project pipeline or an export increase as a complete measure of industrial success.

---

American Economic Senate