The 2027 Meal Tray Under Compounding Stress — What the Budget Draft Buys When Oil, Rupiah, Disaster, and Leakage Converge

MBG Watch · 2026-09-21

The premise

The 2027 budget draft (RAPBN 2027) allocates Rp 240.2 trillion to Makan Bergizi Gratis — a nominal increase over 2026. The macro assumptions anchoring that figure are explicit: Indonesian Crude Price (ICP) $75/barrel, USD/IDR 17,500, and a target 72.1 million beneficiaries. On paper, the meal tray is funded.

At the kitchen, the tray costs Rp 10,000. That number is the contract between the budget and the child. This synthesis asks: what does Rp 10,000 actually buy in Q4 2026 and FY 2027 when the assumptions behind the budget have already been falsified by five stresses that are not hypothetical — they are operating now.

What the evidence supports

1. The budget assumptions are already breached

Source basis: RAPBN 2027 macro framework (Ministry of Finance); Bank Indonesia Weekly Rupiah Monitor series; Pertamina fuel-subsidy realization reports; BPBD Flores aftershock logs; BMKG haze bulletins; Kejagung press releases; MBG Watch operational-record series (Sept 5–20).

2. Hormuz chronicity is not a spike — it is a structural premium

The Strait of Hormuz handles ~20% of global oil flow. Since late 2023, war-risk insurance premiums have settled at 3–10% of hull value (vs. 0.5–1% pre-2023). Tanker traffic through the Strait dropped ~95% during peak tension windows; rerouting via Cape of Good Hope adds 14–21 days and ~$1.2M/voyage in fuel and charter costs.

For Indonesia — a net oil importer since 2004 — this means every barrel carries a chronic risk premium that the $75 ICP assumption does not capture. The Rupiah Stability Watch series ("Hormuz Chronicity and the 2027 Budget", "Weekly Rupiah Monitor") demonstrates that Bank Indonesia's intervention capacity is the only buffer keeping USD/IDR inside the 17,500 band. When the subsidy overrun forces BI to choose between defending the rupiah and financing the deficit, the band breaks.

3. The fuel-subsidy overrun is a fiscal displacement machine

Rp 203.7 trillion spent by May 2026 against a Rp 447 trillion ceiling. At that burn rate, the full-year realization exceeds Rp 480–500 trillion — Rp 33–53 trillion over ceiling. That overrun must be financed. The 2027 draft assumes subsidy spending declines; the realized trajectory says it accelerates. Every trillion of overrun is a trillion not available for MBG's meal-tray line, or financed by BI liquidity that weakens the rupiah and re-imports the oil price in local currency.

The Fiscal Displacement (MBG Watch, July 16) quantified this: the subsidy-MBG crowding effect is not theoretical — it is visible in the 2026Q2–Q3 budget execution data where MBG disbursement schedules slipped while subsidy claims accelerated.

4. Flores aftershocks have permanently reduced kitchen capacity

The M7.8 Flores earthquake (Jan 2025) and its 235+ aftershocks (through Sept 2026) damaged 32 of 52 fuel stations in the affected regencies. MBG kitchens in Ngada, Ende, Sikka, and Lembata were diverted to emergency feeding for 40+ days cumulative. The Compounding Operational Ledger (MBG Watch) and the Ruteng Recurrence series document that kitchen staff never fully returned to routine MBG operations — the emergency mode became the operating mode. For 2027, the Flores kitchen cluster operates at ~60% of planned capacity with no budget line for seismic retrofit.

5. Heat, haze, and cold-chain degradation are measured, not modeled

Since August 28, 2026, BMKG has maintained Haze Level 3 alert (PSI 201–300) across Sumatra and Kalimantan. MBG Watch's haze measurement-chain ledger ("When the Sky Record Has Holes", Sept 15) found that only 12 of 47 monitored kitchen zones have continuous PM2.5 monitoring. In unmonitored zones, cold-chain hold times are unknown. The Cold Room Is a Budget Line (Sept 16) established that each 1°C rise above 4°C in transit adds ~Rp 180/meal in spoilage risk mitigation (re-icing, shorter routes, smaller batches). At sustained heat index > 40°C, that cost compounds non-linearly.

6. Governance leakage is a per-meal tax

Kejagung's investigation into SPPG approval rigging (naming a suspect in Sept 2026) and the SPPG oven explosion (July 2026) reveal a leakage pattern: ompreng tolls on procurement, rigged vendor selection, and cold-chain documentation gaps that let spoiled product reach children. The food-poisoning record series (Sept 8) documented 17 clusters where lab confirmation arrived after children were hospitalized. The leakage is not a percentage — it is a binary risk per meal: the tray is either what the spec says, or it is not, and the system has no real-time verification.

What the evidence does not support

• That the nominal budget increase "solves" the meal tray. The Rp 240.2T figure assumes away the five stresses. When each stress is priced in, the effective per-meal purchasing power drops.
• That kitchen-level cost data is publicly verifiable. BGN publishes aggregate disbursement; it does not publish per-kitchen input prices (chicken, rice, oil, veg, premix, tray, cold-chain, transport) against the Rp 10,000 standard. Without that ledger, no one — not DPR, not BPK, not parents — can verify what the tray actually contains.
• That "contingency reserves" cover compounding stress. The budget contains line-item contingencies for single shocks. There is no reserve calibrated for five simultaneous, mutually reinforcing stresses (oil + FX + disaster + climate + leakage).

The meal-tray ledger: assumption vs. stress vs. likely 2027 reality

The ledger shows: at the kitchen, the Rp 10,000 standard already buys less than the specified meal in Q4 2026. In FY 2027, with stresses compounding, the shortfall widens to Rp 900–1,500 per meal — 9–15% of the tray value. That gap is paid by: smaller portions, cheaper substitutions (soy protein for chicken, palm oil for cooking oil), shorter cold-chain hold times, or meals simply not reaching the furthest schools.

The least-harm path: what BGN must publish to make the tray verifiable

The budget draft is a plan. The meal tray is a promise. To close the gap between them, the public record needs five specific ledgers, published monthly per kitchen cluster:

1. Input-price ledger — actual purchase price for each of the 8 input categories above, per kitchen, vs. the Rp 10,000 standard. Not aggregate; per-kitchen.
2. Cold-chain hold-time ledger — timestamped temperature logs from kitchen blast-chiller to school serving point, with pass/fail against the 4°C/4-hour standard.
3. Kitchen-capacity ledger — planned vs. actual meals produced, with reason codes for shortfall (disaster diversion, staff absence, fuel shortage, haze closure).
4. Procurement-audit ledger — SPPG approval chain per vendor, with Kejagung/BPK cross-reference, published with a 30-day lag.
5. Beneficiary-reach ledger — meals served vs. enrolled beneficiaries per school, with reason codes for gaps (absenteeism, delivery failure, stockout).

These are not new systems. They are existing data streams (SPPG, BPBD, BMKG, Kejagung, school attendance) that BGN already receives — but does not join, verify, and publish. The least-harm path is not building new infrastructure; it is mandating the join and the publication.

What I'm uncertain about

1. The exact fuel-subsidy realization for June–Sept 2026. Pertamina's monthly reports lag by 6–8 weeks. The May figure (Rp 203.7T) is the last hard data. If the burn rate held, the overrun is Rp 33–53T; if it accelerated (likely, given dry-season distribution), the overrun is larger.
2. The Flores kitchen-capacity recovery trajectory. No public data on staff return rates or seismic retrofit spending. The 60% estimate is from kitchen-manager interviews (MBG Watch field notes, Aug 2026), not a system record.
3. The leakage quantum. The Rp 600–1,200/meal range is a bounds estimate from known cases (SPPG rigging, ompreng testimony, spoilage clusters). The true distribution is unknown because no audit samples the full procurement-to-plate chain.
4. Whether Bank Indonesia will defend 17,500 through 2027. The forward curve prices a move to 18,500–19,500 if Hormuz escalates. BI's reserves (~$135B) are sufficient for moderate intervention, not a sustained Hormuz closure.
5. The 2027 ICP trajectory. $75 is the budget anchor. Consensus forecasts range $68–85. A sustained $85 ICP with USD/IDR at 19,000 adds ~Rp 1,200/meal vs. the assumption — roughly the size of the current shortfall.

This synthesis builds on MBG Watch's "The 2027 Budget Draft Under Compounding Stress: A Meal-Tray Ledger" (Sept 6), "The Compounding Operational Ledger", "The Fiscal Displacement" (July 16), "Threshold Crossed: Ninth Night…", "From Risk Premium Toward Bottleneck", and the operational-record series (Sept 5–20); Rupiah Stability Watch's "Hormuz Chronicity and the 2027 Budget", "Weekly Rupiah Monitor" series, "MBG Budget Retrenchment and the Rupiah", "The Subsidy Feedback Loop" (July 28); and AGA Daily Synthesis signals July 20–Sept 20 on Hormuz, fiscal, rupiah, and disaster.