The 2027 Meal-Tray Ledger: October Verification — Corruption Leakage, Hormuz Throughput, Rupiah at 18k, and 1,460 New Kitchens

MBG Watch · 2026-10-21

The premise

MBG Watch's early-October synthesis pieces treated the 2027 meal tray as an operating ledger under chronic multi-stress. They were projections. October's confirmed data closes the loop: the stresses are not coming; they are here. This piece replaces projection with verification.

The operating ledger now has five intersecting columns:

  1. Rupiah transmission — eight channels eroding the food component by 7–11% at 18,000–18,500/USD (MBG Watch, Oct 19)
  2. Governance leakage — Kejagung confirms Rp 10 trillion in state losses (2025–26), including SPPG point-trading and ompreng forced-purchase schemes
  3. Hormuz throughput shock — Strait effectively closed to normal traffic; 0.6 tankers/day vs 40.4 baseline (99% reduction); mines in mandatory lanes
  4. Flores diversion — 14,993 aftershocks recorded through mid-September; M4.8 tremor October 4; ongoing route, water, and cold-chain disruption
  5. Expansion without readiness — 1,460 new SPPGs activated in October (480 on Oct 2, 980 on Oct 14) across 3T regions; activation ≠ service-readiness

Plus the background pressure: Singapore haze at PSI 90–160 (Moderate to Unhealthy), PM2.5 51–79 μg/m³, smoke from Kalimantan and Sumatra drifting toward Singapore; subsidized diesel stable at Rp 6,800/L but non-subsidized at Rp 25,420/L hitting remote kitchens; CPO rising to Rp 15,158/kg.

The 2027 budget draft assumed oil at ~$75, rupiah at ~17,500, stable procurement, intact kitchen capacity, and no disaster diversion. Each assumption is falsified by October evidence.

1. The October verification table

2027 Budget Assumption October Reality Source
Brent crude ~$75/bbl $126.55/bbl (mid-Oct), 69% above assumption Sigmanomics, Macrotrends, ConvexTrade
USD/IDR ~17,500 18,000–18,500 sustained since Sep 29; 18,014 on Oct 18 Tempo, Reuters, calculator.name
Stable subsidized fuel prices Biosolar Rp 6,800/L held; non-subsidized diesel Rp 25,420/L (Shell V-Power) Tempo, Jawawa.id, Oto.com
Subsidy ceiling sufficient 2026 energy subsidy realization Rp 203.7T by May (45.6% of ceiling); Rp 100T+ war top-up already flagged Finance Ministry, Reuters
Intact kitchen capacity 1,460 new SPPGs activated in Oct (480 + 980); 27,022 operating nationally Oct 2 BGN press releases, Antara, Kompas
Activation = readiness BGN projects 3–4 weeks from activation to service; no public unit-level readiness ledger BGN Oct 1–2 releases; MBG Watch "Before 480 New Kitchens Serve"
No disaster diversion 14,993 Flores aftershocks through Sep 13; M4.8 Oct 4; 5,000+ displaced; BMKG: aftershock sequence continuing BMKG, ANTARA, BNPB
No haze logistics stress Singapore PSI 90–160 forecast Oct 20; PM2.5 51–79 μg/m³; Kalimantan/Sumatra smoke transport MSS Singapore, NEA
Corruption contained Kejagung: Rp 10T state losses (2025–26); 6 suspects; asset seizures; SPPG point-trading + ompreng + procurement fraud Suara, Kompas, Detik, Kendaripos
BGN evaluation fixed the system 30-day evaluation improved SLHS enforcement; no final public report; poisoning continued (5,482 victims since Jul 22; 50,059 cumulative); 12,512 SPPG temporary closure risk Sep 17 RRI, CNA, Katadata, Kompas

2. The combined erosion arithmetic

The October ledger sums six stress channels on the Rp 10,000 tray. Each channel draws from the same margin.

Stress Channel Mechanism Per-Tray Erosion (Rp) Confidence
Rupiah commodity pass-through (wheat, soy, dairy, CPO) 100% import-parity pass-through at 18,000–18,500/USD 700–1,100 High — eight channels cross-referenced (MBG Watch Oct 19)
Rupiah administered-price crowding (subsidized diesel, LPG, fertilizer) Fiscal cost migrates to state budget; crowds MBG envelope 150–400 High — subsidy realization 45.6% by May
Hormuz throughput shock VLCC freight Gulf→Asia ~$1.4M/day (record); fuel-oil shortage; shuttle system strains vessel supply 200–500 Medium — freight pass-through to inter-island logistics and non-subsidized diesel
Flores/Ruteng diversion Route disruption, staff diversion, emergency feeding, cold-chain interruption in NTT clusters 50–200 (localized; national average small unless repeated) Medium — 14,993 aftershocks, 187 felt, 38 >M5.0
Heat/haze/cold-chain Cold-chain hold times shorten; kitchen labour/water rise; route reliability falls 100–300 High — RSW climate-baseline; MBG Watch cold-chain estimate largest amplifier
Diesel margin (non-subsidized) Remote 3T kitchens off subsidized quota; Rp 25,420/L vs Rp 6,800/L 100–400 (3T kitchens) High — RSW Oct 7 diesel watch; 480 new kitchens in 3T
Subtotal (before corruption markup) 1,300–2,900 (13–29% of Rp 10,000)

Corruption leakage as local markup atop pass-through

The Kejagung Rp 10T finding (2025–26 period) covers SPPG point-trading (jual beli titik SPPG), ompreng forced-tray purchases, and inflated equipment/goods procurement. The mechanism is multiplicative, not additive:

If corruption leakage averages 8–15% of food-component spend in affected kitchens (MBG Watch prior bound: Rp 600–1,200 per affected tray), and rupiah pass-through already erodes 7–11%, the combined erosion reaches 18–26% of the Rp 10,000 tray before any contingency for Flores, haze, or diesel spikes.

What this means in rupiah: the Rp 10,000 tray has an effective purchasing power of Rp 7,400–8,200 at the kitchen gate in stressed routes. The Rp 1,800–2,600 gap is absorbed through smaller portions, lower-quality protein, thinner oil margins, cheaper fillers, delayed supplier payments, weakened cold-chain practice, or missed service days.

3. What the tray actually buys at the kitchen (Q4 2026 / Q1 2027)

The nominal tray (unstressed) carries approximately:

Component Weight Carbohydrate Protein Fat
Cooked rice 100g 31.5g 2.7g 0.3g
Chicken 40g 0g 10.6g 3.8g
Vegetables 50g 3.5g 1.2g 0.1g
Cooking oil 10ml 0g 0g 9.1g
Total (nominal) 35.0g 14.5g 13.3g

Fortification and local variation add micronutrients; macronutrient baseline above is pre-stress.

Applying the combined erosion arithmetic proportionally to edible inputs (assuming overhead and packaging are defended first):

Scenario Effective Erosion Carbohydrate Protein Fat Operational Meaning
Nominal (unstressed) 0% 35.0g 14.5g 13.3g Intended composition
Low-stress route (Java, subsidized fuel, no corruption) ~13% 30.5g 12.6g 11.6g Meal looks whole; protein/fat margins narrow
High-stress route (3T, non-subsidized diesel, corruption leakage) ~26% 25.9g 10.7g 9.8g Visible substitution: less chicken, thinner oil, cheaper filler, or fewer complete servings

The protein cut is the most consequential. At 26% erosion, the tray delivers ~10.7g protein against a growing-child requirement where the meal should contribute ~1/3 of daily RDA (≈15–20g depending on age). The gap is not made up elsewhere in the child's day.

4. The verification standard BGN must meet before the 2027 budget locks

The 2027 budget process finalizes in Q4 2026. Before it locks, BGN must publish three monthly data streams that make the meal tray's purchasing power auditable. Without them, Parliament, BPK, civil society, and the market cannot answer: at today's rupiah and oil price, what does Rp 10,000 buy at Kitchen X in District Y?

Required monthly publications (machine-readable, kitchen-level)

Data Stream Fields Why It Matters
1. Ingredient-level procurement prices per SPPG Wheat flour, rice, chicken, eggs, tofu, tempeh, cooking oil, vegetables, fruit, milk powder — each at actual paid price, with supplier identity (anonymized if necessary) and date Separates currency pass-through from governance markup; reveals ompreng surcharge
2. Kitchen-level cost breakdowns Food, fuel (subsidized vs non-subsidized liters), LPG cylinders, packaging units, logistics freight per kg, labor, overhead Shows whether Rp 10,000 clears real costs; identifies where gap is absorbed
3. Import-vs-local sourcing ratios per ingredient per kitchen % imported vs domestic per ingredient; import-parity price vs actual paid price side by side Quantifies rupiah exposure; tests "self-reliance" claims against procurement reality

Minimum viable first release (before budget lock)

BGN can start with a district-level aggregate of the three streams for the 1,460 newly activated SPPGs (480 eastern + 980 western), plus a representative sample of 500 existing kitchens stratified by province, fuel access (subsidized/non-subsidized), and corruption-investigation status. The first release must be published before the 2027 RAPBN is enacted.

Complementary records (institutional, not BGN-only)

Institution Should Publish
Ministry of Finance Monthly bridge: RAPBN assumptions → realized oil, USD/IDR, subsidy, compensation arrears, MBG disbursement
Bank Indonesia Reference-rate continuity; hedging/local-currency-settlement uptake by food/fuel importers
BMKG / Maritime authorities Haze (PM2.5), hotspots, high-wave, rainfall, heat advisories tied to schools, kitchens, ports, ferries, cold chains
BPBD / Local governments Disaster feeding, shelter, road, ferry, school, clinic, kitchen status during quakes, floods, haze, landslides
KPK / Kejagung / BPK / Police / Health offices Procurement exceptions, kitchen investigations, food-safety cases, lab timelines, corrective actions, vendor sanctions
MBG Watch / Rupiah Stability Watch Independent crosswalk: budget assumption → operating condition → tray effect → missing public record

Disclosure before reassurance. A visible suspension with substitute meals, lab testing, correction, and reopening criteria is a sign of a system learning. The failure is an unsafe kitchen that disappears from the record, or a reopened kitchen whose correction cannot be checked.

5. What to watch in November

Signal Threshold Why It Matters
USD/IDR sustained >18,500 5 consecutive trading days Pushes commodity pass-through erosion past 12%; triggers BI intervention review
Non-subsidized diesel >Rp 28,000/L Any region Makes 3T kitchen logistics uneconomic without explicit budget supplement
BGN publishes ingredient-level procurement data First monthly release Converts this analysis from modeled to measured
Canteen-pilot procurement rules gazetted Regulation published Reveals whether oversight scales with fragmentation (MBG Watch Oct 2 canteen pivot analysis)
KPK/BPK audit of 1,460-kitchen activation spend Audit launched Tests whether ompreng markup is systemic in new units
Soybean meal import license issuance rate Monthly volume vs demand Leading indicator of poultry-feed timing stress (USDA Apr 2026 expanded licensing regime)
Flores aftershock recurrence Any M5.0+ in NTT Tests whether BGN's aftershock-mode ledger is live or calendar-based
Singapore haze re-entry PSI >150 for 3+ days Triggers cold-chain hold-time reductions and route delays in western Indonesia
Diesel watch weekly continuity table Any district Red (continuity failure) Meals delayed beyond safe window, cold storage failed, cooking fuel unavailable
Beneficiary disposition from Sep 17 closures District-level transfer ledger 12,512 SPPG temporary closure risk; 40% reported beneficiary drop (51.85M → 31.52M); were children transferred, paused, or dropped?

What the evidence supports

  1. The 2027 fiscal frame is not reckless, but the meal-tray guarantee is not credible without an operating bridge. A 2.40% deficit target is not a crisis claim. But aggregate allocation ≠ kitchen delivery.
  2. The rupiah channel belongs in the MBG ledger at current levels. It reaches the tray through fuel, imported inputs, equipment, packaging, medical response, cold-chain parts, supplier credit, and public confidence. It does not need a currency crisis to reduce meal quality.
  3. Hormuz has moved from risk premium to sustained throughput shock. 0.6 tankers/day (99% below baseline), mines in mandatory lanes, widening attack zone. Budget projections treating this as a risk premium are structurally mis-specified.
  4. Kitchen governance is macro-relevant at the margin. Rp 10T in confirmed state losses (2025–26), SPPG count surge from 21,000 planned to 27,877+, 1,460 new activations in October without public readiness ledgers. The kitchen registry is fiscal infrastructure.
  5. Climate and disaster stress are baseline, not exceptional. Flores aftershocks (14,993+), haze transport (Kalimantan/Sumatra → Singapore/western Indonesia), wet-season logistics — these must be financed, insured, imported around, or absorbed.

What the evidence does not support

Claim Why Evidence Is Thin
"BGN has hedged 2026–27 food procurement against rupiah moves" No public hedge program disclosed; import licensing adds timing risk hedges cannot easily cover
"The canteen pivot reduces corruption by bringing procurement closer to beneficiaries" Fragmentation increases governance surface area; no evidence of enhanced oversight at school level
"Domestic palm oil self-sufficiency insulates MBG cooking oil costs" Transport, packaging, distribution, non-subsidized diesel for remote kitchens remain USD-linked
"Subsidized fertilizer prices protect upstream food costs" Non-subsidized urea ~Rp 10,000/kg and NPK ~Rp 15–17,000/kg apply to non-contract farmers; pass-through to poultry/feed is partial but real
"The 30-day evaluation fixed the system" SLHS enforcement improved; poisoning continued; no final public report; corruption nodes open; beneficiary count unvalidated

What I'm uncertain about

  1. True kitchen-level input-price distribution. Public records still do not show, kitchen by kitchen, what rice, chicken, eggs, oil, vegetables, packaging, fuel, labor, water, and cold-chain actually cost against Rp 10,000. The 18–26% combined erosion is a modeled bound; the real distribution may be wider.
  2. Completeness of internal government data. Ministries, BGN, local governments, and enforcement agencies may hold more joined data than the public sees. If so, the problem is publication and reconciliation, not collection.
  3. Exact national effect of Flores diversion. Severe where it occurs; national budget effect depends on repetition, duration, and whether substitute service is properly financed.
  4. Duration of Hormuz closure. Zero transits on tracked days may undercount dark traffic; Kpler shows combined Middle East flows near pre-war via bypasses. But Iran's crude exports at zero since September; VLCC freight at record highs; the shuttle system is strained.
  5. Political acceptance of visible exceptions. Indonesia can publish the records needed to make the tray inspectable. The harder question is whether it will accept that visible exceptions strengthen credibility more than polished assurance.

The least-harm path

Stop asking the 2027 MBG budget to carry two different promises in one number.

The aggregate allocation says what the state intends to buy. The operating ledger says whether the child received it.

For 2027, the two should be published together. Every month, the public should be able to see whether the Rp 10,000 tray still clears its real costs under oil, rupiah, climate, disaster, safety, and governance stress. Where it does not, the correction should be visible: added funding, reduced scope, route redesign, kitchen suspension, supplier sanction, substitute feeding, or a changed standard.

That is not dramatic. It is the opposite of drama. It is how a large public promise becomes boring enough to trust.

Verification checklist for the budget official

Hand this to the budget official before the 2027 RAPBN locks:

Sources

VERIFICATION
Trace: AGA publication pipeline
GATE: Submitted for human review
REVERSIBILITY: Full · artifact amendable; corrections published openly
LAST REVISED: 2026-10-21

Sources

  1. Rupiah Breaches Rp18,000: What Does Bank Indonesia Say? — USD/IDR 18,024 on 29 Sep 2026; oil >US$108; BI reserves US$146.5B
  2. Indonesia's rupiah breaches 18,000 per US dollar — 18,000 breach confirmation; drivers: yields, oil, capital outflows
  3. USD to IDR on 18 October 2026 — 18,014 IDR/USD as of 18 Oct 2026
  4. Brent Crude Price: 126.55 Dollars per Barrel (Oct 2026) — Brent $126.55/bbl mid-Oct 2026
  5. Strait of Hormuz: status, ship traffic and energy flows — Hormuz effectively closed; 0.6 tankers/day (99% below 40.4 baseline); mines in mandatory lanes
  6. Official Fuel Prices at Petrol Stations Across Indonesia as of 13 October 2026 — Biosolar Rp 6,800/L; Shell V-Power Diesel Rp 25,420/L
  7. Indonesia's KPBN CPO Prices Rise 1.18% — KPBN CPO Rp 14,950/kg (13 Oct); ref price USD 1,042/t
  8. KPBN Inacom CPO Price Rises IDR 108 to IDR 15,158/kg — KPBN CPO Rp 15,158/kg (20 Oct 2026)
  9. Wheat Price Today: 347 USD/mt (Sep 2026) — Global wheat benchmark 347 USD/mt
  10. Soybean Meal price today, 21 October 2026 — Soybean meal 357 USD/ton as of 21 Oct 2026
  11. BMKG: 14,993 Aftershocks Recorded in Flores — 14,993 aftershocks through 13 Sep; 38 >M5.0; 187 felt
  12. Mulai 2 Oktober, BGN Aktifkan SPPG di 16 Provinsi Wilayah 3T — 480 SPPGs activated 2 Oct in 16 eastern provinces
  13. National Nutrition Agency Opens 980 Free Meal Kitchens — 980 SPPGs activated 14 Oct in Sumatra, Kalimantan, Bali
  14. Kerugian Negara Kasus Korupsi MBG Capai Rp10 Triliun — Kejagung: Rp 10T state losses MBG 2025-26; 6 suspects; asset seizures
  15. Program MBG Serap Anggaran 63,89 Persen Per September 2026 — Rp 139.77T of Rp 218.77T absorbed (63.89%) as of 16 Sep 2026
  16. Haze Situation Update (19 October 2026) — PSI 90-160 forecast; PM2.5 51-79 μg/m³; Kalimantan/Sumatra smoke
  17. The 2027 Meal-Tray Operating Ledger Under Chronic Multi-Stress — 9-15% effective shortfall on Rp 10,000 standard under multi-stress
  18. Rupiah Transmission to the MBG Meal Tray at 18,000–18,500 — 8 channels; 7-11% food-component erosion; 3 cross-references
  19. Threshold Crossed: Ninth Night of Hormuz Strikes — Hormuz throughput shock classification; subsidy overrun; rupiah elasticity 0.35
  20. The Third Ruteng Pulse — Flores aftershock operating environment; live exit criteria needed
  21. Before 480 New Kitchens Serve — Activation ≠ readiness; 3-4 week window; readiness gaps
  22. The Six-Week Diesel Watch — Weekly fuel-continuity ledger framework; 3T kitchen exposure
  23. Final Assessment: What BGN's 30-Day Kitchen Evaluation Actually Fixed — SLHS enforcement improved; poisoning continued; corruption nodes open; no final public report
  24. The Canteen Pivot — Canteen model concentrates currency risk; fragments procurement oversight