When the Budget and the Grocery Bill Move at Once: MBG Under the Rupiah and the Oil Shock

MBG Watch · 2026-07-03

The premise

Two numbers moved in opposite directions this spring, and a child's plate sits between them.

On one side, the budget for Indonesia's Free Nutritious Meals program — Makan Bergizi Gratis, or MBG — was cut from Rp 335 trillion to Rp 268 trillion in the 2026 state budget, a reduction of Rp 67 trillion, almost exactly 20 percent (ANTARA, Jakarta Globe, May 2026). On the other side, the price of the food that money buys kept rising: Indonesia's headline inflation reached 3.08 percent year-on-year in May 2026, led by food — chili, cooking oil, shallots — and by energy (BPS, via Jakarta Globe and The Jakarta Post, June 2, 2026). The rupiah, meanwhile, had fallen from roughly 14,670 to 17,950 per US dollar, a depreciation of about 22 percent, raising the landed cost of imported food staples and inputs (Rupiah Stability Watch, Weekly Rupiah Monitor, June 10, 2026).

MBG is meant to reach 83 million children and pregnant women. The question this piece asks is narrow and concrete: when the program's money shrinks and the price of food rises at the same time, does a child still get the meal that was promised — and does the household around that child come out ahead?

The honest answer is layered. The squeeze is real, but it is smaller than "budget down 20 percent" implies, it lands hardest on the places already least served, and the most important risk is not the size of the cut but a number that has not moved at all.

What the evidence supports

The per-meal budget has been frozen in nominal terms — and that is where the real erosion lives. Throughout 2025 and into 2026, the National Nutrition Agency (BGN) held the allocation at Rp 10,000 per portion for raw materials across most of the country, with a total of about Rp 15,000 per child per day once production and distribution are included (VOI; Tempo; Katadata, 2025–2026). BGN's leadership stated explicitly that the standard would not rise even as the overall budget envelope grew, because the growth reflected more beneficiaries, not a richer meal. A fixed Rp 10,000 against food inflation of roughly 3 percent means the same coin buys a thinner plate each quarter. This is the quiet mechanism that matters more than the headline cut: nutrition per meal can decline without a single line item being formally reduced.

The headline 20 percent cut is partly an accounting correction, not pure austerity. Actual MBG spending reached only about Rp 75 trillion ($4.24 billion) in the year to April 30, against allocations of Rp 268–335 trillion (Finance Minister Purbaya Yudhi Sadewa, via The Diplomat, June 22, 2026). The program has been unable to spend what it was given, largely because the kitchens that prepare the meals were built more slowly than planned (Tempo). Much of the Rp 67 trillion removed was money the program was not going to disburse this year. The cut therefore tightens the future ceiling more than it removes meals already on tables — an important distinction the topline number obscures.

The fiscal pressure and the household pressure do share a root. The government's drive for "budget efficiency" intensified after the closure of the Strait of Hormuz in March 2026 drove global oil prices up, costing billions in fuel subsidies (The Diplomat, June 22, 2026), and against the backdrop of the rupiah's slide. Those same external shocks — a weaker rupiah, costlier imported energy and food — are what eroded household purchasing power. A sister assessment finds the bottom-40 percent of households lost about 3.25 percent of real income to depreciation-driven food prices (Rupiah Stability Watch, Essential Goods Impact Assessment, June 2026). So the program's belt-tightening and the family's grocery bill are, in this limited sense, two effects of one cause.

The program is adjusting in ways that have an equity logic — and an equity risk. BGN has suspended meals during school holidays (June 22–July 13), saving over Rp 3 trillion; defunded around 39,000 students in 76 schools judged economically capable, reallocating capacity to remote areas; and floated reducing service to four days a week to save roughly Rp 50 trillion a year (The Diplomat; idnfinancials, 2026). Targeting the better-off out and the underserved in is defensible on paper. But it raises the second-order question of whether the remote areas now slated to receive that capacity have the kitchens and cold chains to use it — the same infrastructure gaps our earlier Food Safety Crisis in MBG Rural Rollout documented.

What the evidence does not support

It does not support the claim that the budget cut was caused by the currency crisis. The cut is better explained by low budget absorption, an efficiency-and-corruption response — former BGN head Dadan Hindayana and two deputies were arrested this month over alleged state losses (The Diplomat, June 22, 2026) — and broad fiscal caution after the oil shock. The rupiah is part of the fiscal backdrop, not the documented trigger. Calling the cut and the household squeeze "two faces of the same crisis" is half right: they share external causes, but the cut is not a transmission of the currency crisis into the program.

It does not support a claim that MBG, on its own, leaves a beneficiary household worse off. Consider the arithmetic plainly, and treat it as illustrative, not measured. A child's MBG meal funded at roughly Rp 15,000 per school day, over about 20 school days a month, is worth on the order of Rp 300,000 a month in food the household would otherwise buy. For a bottom-40 percent household, that is a meaningful transfer — plausibly several percent of monthly food spending. The 3.25 percent real-income loss from food inflation is a genuine headwind, but for a household that actually receives the meals on most school days, the MBG transfer is the larger of the two effects. The net direction is most likely still positive. The caveat is decisive: this holds only where the meals are reliably delivered, which is precisely what is least certain in the places that need it most.

It does not support either alarm or reassurance about program collapse. There is no evidence the program is failing at current funding; there is also no evidence the frozen per-meal rate is sustainable if food inflation persists. Both the "scrap it" demand voiced at recent Jakarta protests and the official "still adequate" line outrun what the data can carry.

The least-harm path

The upstream cause here is not the size of the budget; it is a per-meal rate that does not move while food prices do. The lowest-harm, most reversible adjustment is to index the per-portion allocation to regional food prices — explicitly already acknowledged in principle, since Papua and Maluku are funded at up to Rp 30,000 against Java's Rp 10,000 (Bisnis, 2025). Extending that logic to a quarterly inflation adjustment would protect the meal's nutritional content without re-inflating the topline budget, because the binding constraint is real value per plate, not total rupiah allocated.

Three further steps follow the same proportional, reversible principle:

None of these requires spending more. They require spending the same money against a moving target instead of a frozen one.

What I'm uncertain about

In order of how much it matters:

  1. Delivery reliability in the places now prioritized. The whole net-benefit case rests on meals actually arriving in remote and outer-island schools. We have strong evidence of infrastructure gaps there and weak evidence on current delivery rates. This is the gap that could flip the conclusion.
  2. The real value of a Rp 10,000 meal today. We know the nominal rate and the inflation rate; we do not have a verified measurement of the nutritional content actually being served per portion in mid-2026, or how much it has thinned since the rate was set.
  3. Household substitution behavior. Whether families receiving MBG redirect saved food money toward other nutrition or toward non-food needs determines whether the meal's benefit compounds or merely offsets. We have no Indonesian data on this for 2026.
  4. The durability of the external shocks. If the Hormuz disruption eases and the rupiah stabilizes, the fiscal and household pressures both relax, and the frozen rate becomes a manageable problem. If they deepen, the rate freeze becomes the program's central vulnerability.

The headline says the budget fell by a fifth. The plate tells a quieter, more important story: it was promised Rp 10,000, and Rp 10,000 buys less than it did.

The threshold to watch is not a budget figure. It is the first credible regional measurement showing the real nutritional value of a delivered MBG meal has fallen below the standard the program set for itself. Until that number is published, no one — defender or critic — can honestly say whether the meal still does what it was built to do.

Sources: BPS via Jakarta Globe and The Jakarta Post (June 2026); The Diplomat (June 22, 2026); ANTARA and Jakarta Globe (May 2026); Tempo, VOI, Katadata, Bisnis (2025–2026); idnfinancials (April 2026); Rupiah Stability Watch, Essential Goods Impact Assessment and Weekly Rupiah Monitor (June 2026). Builds on MBG Watch, Early Rollout Evidence Review and Food Safety Crisis in MBG Rural Rollout. Household-level figures are illustrative arithmetic from cited rates, not measured outcomes; flagged as such in the text.