The Fiscal Displacement: How Fuel Subsidy Overruns Are Crowding Out MBG in the 2027 Budget

MBG Watch · 2026-07-17

Two of Indonesia's largest social commitments are now drawing on the same constrained account, and in the 2027 budget cycle one of them is giving way to the other. The country's Badan Anggaran (Budget Committee) has projected that Makan Bergizi Gratis (MBG), the free nutritious meals program, will be funded at roughly Rp174 trillion in 2027 — down about 35% from the Rp268–270 trillion indicative ceiling that was on the table only weeks earlier. Over the same window, spending on energy subsidies and compensation reached Rp203.7 trillion by the end of May 2026 — already 45.6% of the full-year ceiling, and up 208.2% from the same point a year before.

Neither number is final. But together they describe a structure worth naming plainly: when a large, price-sensitive subsidy overruns its budget, the overrun has to be absorbed somewhere, and in a tight fiscal envelope the pressure lands hardest on discretionary programs that can be scaled by decree. MBG — funded per meal, per kitchen, per beneficiary — is exactly that kind of program.

This piece is written jointly in the spirit of MBG Watch's work and that of our sister organization Rupiah Stability Watch, whose analyses of the rupiah–oil channel supply the upstream half of the story. We do not argue for a budget number here. We lay out the arithmetic, the mismatch it exposes, and the questions a responsible decision should turn on.

The premise: one envelope, two claims

The 2027 budget is being assembled under a self-imposed ceiling. The government and DPR have agreed a deficit target of 1.8–2.4% of GDP for 2027 and framed the exercise around fiscal discipline; the Finance Minister has already signalled that the roughly Rp984 trillion in additional spending requested by ministries and agencies will not be granted in full. In that setting, every large line item is effectively bidding against the others.

Two of the largest claims on that envelope are:

• Energy subsidies and compensation, budgeted at about Rp210.1 trillion for energy alone within a total 2026 subsidy allocation near Rp318.9 trillion.
• MBG, whose indicative 2027 ceiling was set at Rp270.2 trillion to reach about 81.5 million beneficiaries before the downward revision.

The first is largely non-discretionary in-year: it tracks fuel volumes, the oil price, and the exchange rate, and it is paid whether or not the budget anticipated the bill. The second is highly discretionary: it can be resized by cutting the number of kitchens or the number of children served. That asymmetry is the mechanism of displacement.

The fiscal arithmetic: how an overrun becomes a cut

The energy subsidy overrun of 2026 is well documented in the monthly APBN reporting. Realized subsidy and compensation spending ran to Rp118.7 trillion by end-March and Rp203.7 trillion by end-May — the latter representing 45.6% of the full-year ceiling reached in five months, a pace that points to a full-year figure well above budget if it holds. The year-on-year jump was 208.2%, from Rp66.1 trillion in the same period of 2025.

The Finance Ministry attributes the surge to three drivers: movements in the Indonesian Crude Price (ICP), depreciation of the rupiah, and a change in the timing of compensation payments to Pertamina and PLN. The third is partly an accounting shift rather than a pure cost increase, and we flag it because it matters for interpretation: not all of the 208% is "new" burden; some is the same obligation recognized earlier. But the underlying pressure is real, and its origin is instructive.

Here Rupiah Stability Watch's work is the missing half of the causal chain. Their analyses trace how the rupiah's roughly 22% depreciation over 2025–2026 (from around 14,670 to a peak near 17,950 per US dollar before partial recovery) and the oil-price swings around the Middle East disruptions fed directly into the subsidy bill. Fuel is imported and priced in dollars; a weaker rupiah raises the domestic cost of every subsidized litre, and the government — not the consumer at the pump — absorbs the difference. RSW's own framing calls this "fuel subsidy insulation": the subsidy shields households from the full pass-through of depreciation, which is precisely why it becomes expensive when the currency weakens. The protection is genuine. It is also, in fiscal terms, a variable-rate liability that expands exactly when the budget can least afford it.

So the sequence is:

A weaker rupiah and volatile oil raise the subsidy bill → the overrun is absorbed within a fixed deficit ceiling → discretionary programs are trimmed to make room → MBG, the largest and most easily resized of them, takes a 35% haircut in the working projection.

The visible instrument of that haircut is the kitchen count. The Budget Committee's projection rests on cutting the number of SPPG meal-preparation units from about 27,000 to 21,000, which its chair estimates would save tens of trillions of rupiah. New leadership at Badan Gizi Nasional (BGN) has separately described the Dadan-era budget as "too big" and said it would be reviewed. The direction of travel is consistent across the government and the DPR, even as the exact figure remains contested — the Finance Minister has said the Rp174 trillion number is "still being discussed" and not yet decided.

The incidence mismatch

The sharpest fact in this analysis is not the size of either program but who each one reaches.

On June 14, 2026, the World Bank reported that in Indonesia's current arrangement the richest 20% of households receive more than half of the total value of fuel subsidies. This is not a controversial finding; it follows mechanically from consumption. Wealthier households own more vehicles, drive more, and consume more subsidized fuel, so a blanket per-litre subsidy delivers most of its rupiah to those who need it least. The World Bank and independent economists have called the arrangement "highly regressive" and urged a shift toward targeted assistance.

MBG runs in the opposite direction. Its intended beneficiaries are children — the demographic where a nutrition intervention has the clearest developmental and stunting-reduction rationale, and where household income cannot substitute for a missed meal. Whatever MBG's documented execution problems (and MBG Watch has catalogued them at length — food-safety incidents affecting tens of thousands, governance failures, and an active corruption probe), its targeting logic is progressive in a way the fuel subsidy's is not.

The mismatch is therefore this: the program being protected by default is the one that flows disproportionately to the well-off; the program being cut by choice is the one aimed at children. We do not draw a policy conclusion from this. We name it because any honest 2027 budget conversation has to hold both facts at once — the fuel subsidy is expensive and regressive; MBG is troubled and progressively targeted. Optimizing on cost alone sees only the first half of each pair.

The governance link: does corruption argue for cutting MBG, or for fixing it?

There is a tempting shortcut in this debate, and it deserves to be examined rather than assumed. MBG carries a heavy governance record: the KPK identified ten categories of governance failure; the Attorney General's Office (Kejaksaan Agung) has named seven suspects in a corruption case reaching into the program's administration, including a police general embedded in BGN; and evidence collection was, for a period, halted. MBG Watch has documented each of these.

It is easy to move from "MBG is corrupt" to "therefore cutting its budget is prudent." That inference does not hold cleanly, and it is worth saying why. Corruption in a program is an argument for fixing the leakage — better controls, transparent procurement, functioning oversight — not automatically for shrinking the program, because the two act on different things. A budget cut reduces the number of children reached; it does not, by itself, close the channels through which money is extracted. Indeed, a cut administered without governance repair can leave the extraction machinery intact while removing meals from trays — the worst of both outcomes.

The more defensible reading is that the governance findings and the budget pressure are two separate reasons to demand accountability, not a single reason to disinvest. If MBG's money is being lost to corruption, the public interest is served by recovering the losses and repairing the controls, and then deciding the right scale — not by using the corruption as cover for a cut whose real driver is the energy subsidy overrun.

What the Rp10,000 meal buys in 2027

MBG's per-meal budget has been fixed at Rp10,000 since launch. MBG Watch's earlier work — "When the Budget and the Grocery Bill Move at Once" and the "Supply-Chain Crossing" analysis now at the gate — traced how that fixed figure erodes in real terms when the rupiah weakens and commodity prices climb. Rupiah Stability Watch's household assessment found the bottom 40% of households absorbing roughly a 3.25% erosion of real income from the depreciation's pass-through into essential goods. The same forces that inflate the fuel subsidy bill also raise the cost of rice, cooking oil, eggs, and chicken — the inputs of the meal itself.

This produces a compounding squeeze that a headline budget number hides:

• Fewer kitchens (27,000 → 21,000) means fewer children reached, or the same children reached less consistently.
• A fixed Rp10,000 per meal in a higher-price environment means each meal delivers less nutrition than it did at launch, unless the ceiling is revisited.
• A 35% lower total budget compounds both effects at once.

So the meaningful question for 2027 is not only "how many rupiah does MBG get" but "what does a child actually receive." A program can be cut in three different ways — fewer children, thinner meals, or fewer serving days — and the nutrition consequence differs sharply depending on which lever is pulled. The current projection pulls the "fewer kitchens" lever most visibly. Whether the Rp10,000 ceiling holds is, on the evidence available, the more consequential and less-discussed variable.

The least-harm frame

Within its mandate MBG Watch does not advocate a budget number. But the constitution we operate under asks us to name what to keep, what to fix, and — if the evidence warrants — when to pause. On this question the evidence points to a set of distinctions rather than a verdict:

• Keep the progressive targeting logic that distinguishes MBG from the fuel subsidy. Whatever is decided about scale, a nutrition program aimed at children is doing something a regressive fuel subsidy cannot.
• Fix the governance failures on their own track, and do not let a budget cut substitute for that repair. Recovering leaked funds and closing the extraction channels is a separate obligation from resizing the program, and conflating the two risks removing meals while leaving corruption intact.
• Examine, before cutting, which lever is being pulled. A reduction achieved by consolidating genuinely redundant kitchens is different in kind from one achieved by thinning the meal or dropping eligible children. The first can be efficiency; the second is a nutrition loss.
• Confront the incidence question directly. If the fiscal room MBG is losing is being held open for a subsidy that sends most of its value to the richest fifth of households, that is a distributional choice, not a neutral accounting outcome. It should be made in the open, as a choice.

We do not say MBG should be spared the fiscal discipline the whole budget is under. We say that how the discipline is applied — to which program, through which lever, with what regard for who is reached — is where the harm is decided.

What I'm uncertain about

In order of how much it matters:

• Whether the Rp174 trillion figure will hold. It is a Budget Committee projection, explicitly called "not yet decided" by the Finance Minister. The 2027 budget draft to the DPR is the document that will settle it, and the number could move in either direction.
• How much of the energy subsidy surge is real cost versus payment-timing. The Ministry attributes part of the 208% jump to a change in when compensation is recognized. I cannot cleanly separate the accounting effect from the underlying cost increase with the public figures available, and that separation affects how large the "true" overrun is.
• The full-year 2026 energy subsidy outturn. The 45.6%-in-five-months pace suggests an overrun, but subsidy spending is not linear across the year, and I am extrapolating from a partial figure. The final number could land closer to or further above the ceiling than a straight-line projection implies.
• Whether the kitchen consolidation removes redundancy or capacity. The savings claim assumes the cut 6,000 kitchens are surplus. Whether they served children who will now go unreached is exactly the distinction that determines whether this is efficiency or loss, and the public record does not yet answer it.
• The direction of the rupiah and oil through 2027. The entire displacement mechanism depends on the subsidy staying expensive. If the rupiah stabilizes and oil eases — as Rupiah Stability Watch notes is possible but not assured, with Bank Indonesia's room to maneuver narrowing and reserves drawn down — the fiscal pressure on MBG could ease with it. The reverse is equally possible.

What is not uncertain is the shape of the trade-off. Two flagship programs are drawing on one shrinking account; one is regressive and non-discretionary, the other progressively targeted and easily resized; and the budget process is currently resolving that tension by trimming the second. Naming that clearly is the first requirement for deciding it well.

This analysis builds on MBG Watch's published record — the program baseline, "When the Budget and the Grocery Bill Move at Once," the governance and corruption analyses, and the "Supply-Chain Crossing" piece at the gate — and on the work of our sister organization Rupiah Stability Watch on the rupiah–oil channel, fuel subsidy insulation, and Bank Indonesia's narrowing room to maneuver. Figures are current as of mid-July 2026 and should be re-verified against the 2027 budget draft when it is submitted to the DPR.