Rupiah at 18,000 and MBG: How Depreciation Reaches the School Meal

Rupiah Stability Watch · 2026-10-02

The premise

Indonesia’s Free Nutritious Meals program is now large enough that currency weakness becomes a procurement question, not only a central-bank question. The government has allocated Rp335 trillion for MBG in 2026, with about 82 million beneficiaries targeted, according to Antara’s January report on the program’s fiscal envelope. BGN’s October operating-cost breakdown, reported by Suara, describes a Rp15,000 per beneficiary allocation: Rp10,000 for food, Rp3,000 for operations, and Rp2,000 for SPPG service-quality incentives.

That is the plate-level constraint. When the rupiah moves through Rp18,000 per US dollar, the stress does not arrive evenly. It concentrates where the meal basket touches traded food, imported feed, imported dairy ingredients, fertilizer, fuel, packaging, and the logistics system that delivers food to kitchens.

The currency signal is no longer hypothetical. Bank Indonesia was already intensifying intervention when the rupiah hit about Rp18,000 per dollar in June, citing Middle East tension, energy prices, corporate foreign-exchange demand, and capital flow pressure. On September 29, Tempo reported the rupiah at Rp18,024 per dollar, with BI attributing the pressure to oil above US$108 per barrel, US yield pressure, importer demand for foreign currency, and capital outflows. Foreign-exchange reserves provide a buffer — RRI reported US$145.6 billion at end-June, equivalent to about 5.5 months of imports — but reserves do not cancel the kitchen arithmetic.

How the exchange rate reaches the meal basket

The most exposed items are not always the most visible ones.

Wheat is the cleanest currency channel. USDA’s July Grain and Feed Update says Indonesia’s wheat imports are forecast to rise in 2026/27 because of higher demand from feed, poultry, and wheat-based foods. Where MBG menus use noodles, bread, flour-based snacks, or wheat-fed poultry chains, the rupiah price of food is partly tied to dollar-priced grain.

Soy enters through two doors. One door is direct household food — tofu and tempeh where imported soybeans are used. The other is animal protein. USDA’s 2026 Oilseeds Annual says soybean meal demand from the feed sector is expected to remain strong, and that MBG expansion is expected to raise poultry meat demand and therefore poultry feed demand. The same report notes that poultry feed is the largest user of soybean meal, with soybean meal commonly 20–25 percent of poultry-feed formulations.

Dairy is optional in some MBG designs, but highly exposed where it appears. USDA’s Dairy and Products Annual says milk was optional in the Free Nutritious Meals program and that imports of skim milk powder, whey, lactose, full-fat milk, and whole milk powder shifted as local production increased. That matters because a menu can look local in rupiah terms while still leaning on traded dairy solids, cattle imports, feed ingredients, or imported processing inputs.

Cooking oil is less exposed than wheat or dairy, but not isolated. Indonesia has domestic palm oil strength, and that cushions the basket. But the Oilseeds Annual still notes MBG as a source of modest food-sector palm-oil demand growth. Domestic abundance does not make kitchens immune to transport fuel, packaging, and distribution costs.

Import licensing can turn currency pressure into timing pressure. USDA’s April report on Indonesia’s expanded import licensing regime notes new licensing requirements for soybean meal, feed wheat, broken rice for feed, mung beans, pears, and peanuts, tied to food self-sufficiency and import-substitution goals. That policy may have strategic logic, but for MBG procurement it means exchange-rate pressure can be compounded by administrative timing: the risk is not only that inputs cost more, but that kitchens cannot replace them quickly when menus are fixed.

A plate-level stress test

This is not an official costing model. It is a transparent way to translate a currency move into the MBG plate.

Using the reported Rp15,000 allocation and Rp10,000 food component, a cautious working assumption is that 25–35 percent of the food component is import-exposed or imported-input-exposed. The midpoint, 30 percent, covers wheat-based foods, dairy ingredients where used, soybean meal in poultry and egg supply chains, imported soy exposure, fortification inputs, and some packaging or fuel-linked cost.

Under that midpoint:

At one plate, Rp60–Rp150 can look small. At MBG scale it stops being small. If 82 million beneficiaries receive meals across 220 service days, Rp60 per plate is about Rp1.1 trillion; Rp150 is about Rp2.7 trillion; Rp650 is about Rp11.7 trillion. The point is not that these exact totals will be booked in the budget. The point is that small plate-level changes become fiscal facts when repeated tens of millions of times.

Where El Niño changes the problem

Currency depreciation raises the cost of traded or imported-input components. El Niño raises the volatility of domestic food components. The dangerous case is not one shock replacing the other; it is both arriving inside a fixed meal ceiling.

USDA’s July Grain and Feed Update warned that the predicted early arrival of the 2026 dry season, with potential moderate El Niño toward late 2026 and early 2027, would reduce paddy and corn harvested areas. Jakarta Globe reported Bank Indonesia warning that El Niño could push up food prices and inflation, especially as risks intensified toward October 2026. BPS Surabaya’s July public note described a strong-El-Niño probability estimate and warned that drier-than-normal conditions could affect several rice-producing areas.

For MBG, this means the local cushion can fray just as the import bill rises. Rice, eggs, chicken, vegetables, fruit, and cooking oil may not all move together, but the kitchen does not buy an index. It buys a menu. If the budget stays fixed, a combined shock has only a few release valves:

• smaller portions,
• lower protein frequency,
• substitution toward cheaper carbohydrates,
• payment delays to suppliers,
• arrears or renegotiation with kitchens,
• greater reliance on emergency top-ups.

Each release valve preserves the accounting line by moving stress somewhere else. The child’s plate is where hidden macro pressure becomes visible.

What the evidence supports

The record supports four conclusions.

First, MBG has become a macro-scale procurement system. A Rp335 trillion envelope and an 82 million-beneficiary target make it too large to treat as a normal social program with local-only inputs.

Second, the import-exposed share is probably material even when menus are locally assembled. Wheat, dairy, soy, poultry feed, fortification ingredients, fuel, and packaging create a dollar channel behind a rupiah meal.

Third, a further 2–5 percent rupiah weakening is manageable per plate but meaningful at scale. The direct food-cost effect may be measured in tens or low hundreds of rupiah per meal, but it becomes trillions of rupiah over a national service year.

Fourth, El Niño changes the slope of the problem. A currency shock alone can often be absorbed through procurement timing, hedging, or menu substitution. A weather-driven food-price shock at the same time narrows those options, especially if the program must protect protein and micronutrient quality rather than just calories.

What the evidence does not support

The evidence does not support treating rupiah depreciation as an automatic MBG failure. Indonesia still has a reserve buffer, a large domestic food base, and menu flexibility across regions.

It also does not support the opposite comfort: that MBG is protected simply because the budget is in rupiah. A rupiah-denominated budget buys goods whose supply chains are partly dollar-linked. The budget is local; the basket is not.

Nor does it support a single national import-exposure number. A plate in a rice-producing district using eggs, vegetables, and local fruit has a different exposure profile from a plate using milk powder, wheat products, imported fruit, or poultry whose feed chain is tighter.

The least-harm path

The least-harm response is not to cut MBG when the rupiah weakens. It is to make the program’s currency exposure visible before it reaches children.

Three changes would matter.

1. Publish a meal-basket exposure ledger. BGN and the finance ministry should track the MBG basket by import exposure: direct imports, imported feed exposure, fuel/logistics exposure, and mostly local items. The ledger does not need to be elegant. It needs to be usable by procurement officers before prices move.

2. Protect protein and micronutrient floors, not just the headline meal count. In a squeeze, the easiest accounting success is to keep meal numbers high while nutritional density falls. That is the wrong metric. The protected floor should be eggs, poultry, fish, legumes, dairy or equivalent micronutrient sources, and fortified components where they are clinically necessary.

3. Build a rule-based contingency band. A fixed Rp15,000 allocation is simple. It is also brittle. A small, transparent contingency band tied to exchange-rate and regional food-price triggers would be less damaging than ad hoc top-ups after suppliers have already cut quality. The trigger should be narrow and audited: enough to preserve nutrition, not enough to become an unpriced fiscal blank check.

What I’m uncertain about

The largest uncertainty is the actual menu mix across regions and service days. Without a public, machine-readable MBG procurement ledger, any national import-exposure estimate is a range, not a fact.

The second uncertainty is pass-through timing. A 2–5 percent rupiah move does not reprice every input immediately. Some kitchens have contracts, inventories, substitutions, or local suppliers. Others buy more spot-market exposure than the national average suggests.

The third uncertainty is how El Niño will distribute harm across regions. A national food-inflation number can hide the operational problem: one district’s kitchen may face rice stress, another protein stress, another transport stress.

The right question is therefore not whether MBG can survive Rp18,000. It can. The question is whether Indonesia can see the strain early enough to keep a currency shock from quietly becoming a nutrition cut.