What BGN Published (and Didn't) During the October 14 Haze Event

MBG Watch · 2026-10-24

The premise

On 14 October 2026, Singapore's 24-hour PSI hit PM2.5 288 µg/m³ — the highest since the 2015 crisis — and 12% of schools closed (Channel News Asia, 14 Oct 2026). The same haze plume, driven by forest and land fires (karhutla) across Sumatra and Kalimantan, pushed air quality in Pekanbaru, Palembang, Jambi, and Pontianak into "hazardous" bands (ISPU > 300 in Jambi on 5 Oct; Pekanbaru "zero visibility" episodes from 8 Oct). Indonesia's border regions — Riau, Jambi, South Sumatra, West Kalimantan — were already in Pembelajaran Jarak Jauh (PJJ).

MBG Watch's Continuous Monitoring Framework (22 June 2026) treats an unhealthy-haze threshold crossing as an operating-status test: when haze reaches unhealthy levels, BGN should publish a single, consolidated record covering six domains — kitchens, routes, schools, care routes, worker protection, cooked batches — so the public, auditors, and the program itself can see what actually happened to the meal tray.

On 23 October, MBG Watch published "Haze at the School Door: The Operating-Status Record MBG Needs Now" demanding exactly that record. This briefing answers the accountability question: what did BGN actually publish (or not publish) between 14 and 23 October 2026?

What the evidence supports

1. BGN's national-level publications (14–16 Oct)

All three are reactive, procedural, and kitchen-centric — they address whether to stop, not how the operating status looks across the six domains.

2. Regional halts reported by media (not BGN channels)

These are media reports of regional actions, not BGN-published operating-status records. They are fragmentary, region-by-region, and lack the six-domain structure.

3. Singapore threshold context (external but decisive)

• 14 Oct: PM2.5 288 µg/m³ (west Singapore, 15:00), 12% school closure rate (CNA, 14 Oct).
• 14 Oct: Timor-Leste closed all schools nationally for three days (Reuters, 14 Oct).
• 15 Oct: BMKG satellite showed hotspots across Kalimantan, Sumatra, Papua; wind transported haze to Singapore and Malaysia (Liputan6, 12 Oct; CNBC Indonesia, 15 Oct).
• 15 Oct: 1,813 domestic flights disrupted to date (Kompas Travel, 15 Oct).

What the evidence does not support

1. No consolidated operating-status record exists. BGN did not publish a single document (or linked set) covering: kitchen status (paused/rerouted/substituting/continuing), route status, school-level service status, care-route status, worker protection measures, or cooked-batch disposition — for any region, let alone nationally.

2. No haze-specific notices on Radar MBG. The public portal radar-mbg.bgn.go.id shows daily menus. A manual check on 24 Oct reveals no banner, alert, or filter for "haze operations," "PJJ mode," or "operating status" for the 14–23 Oct window.

3. No haze-specific notices on SDMO. The internal Biro SDMO portal (birosdmo.bgn.go.id) is login-gated; no public-facing haze operating notices were found.

4. No cross-regional kitchen/route/school ledger. The Meranti halt (since 9 Oct), Sumut-Aceh halt (16–21 Oct), and Mamuju partial halt (20 Oct) were never aggregated into a single BGN view. A decision-maker could not see the national picture.

5. No worker protection protocol published. The Oct 16 mechanism mentions "accountable distribution" of already-cooked meals but does not specify PPE, exposure limits, transport safety, or kitchen ventilation standards for haze conditions.

6. No complaint-channel haze tagging. BGN's complaint channels (whistleblower, call centre, Radar MBG feedback) carried no haze-specific tag or reporting category in the Oct 14–23 window.

The least-harm path

The Continuous Monitoring Framework (CMF) defines four tiers:

Classification: Elevated Concern.

Why not Routine? The Singapore PM2.5 288 µg/m³ threshold was crossed. PJJ was active in at least four provinces (Aceh, North Sumatra, Riau, Jambi, South Sumatra, West Sulawesi) simultaneously. The CMF treats this as a threshold crossing, not business as usual.

Why not Urgent Investigation? No verified evidence yet of missed meals causing acute malnutrition, food safety incidents from haze-exposed batches, or worker health events. The gap is transparency, not (yet) outcome harm.

Why not Program-Viability Review? This is a single event window. The pattern would need repetition across multiple haze seasons or convergence with other stress channels (diesel, rupiah, corruption leakage) — which the joint Rupiah Stability Watch / MBG Watch synthesis (23 Oct) notes is beginning to appear, but not yet met.

What I'm uncertain about

1. Exact count of affected SPPGs nationally. Media reports cover Sumut-Aceh, Meranti, Mamuju. Riau (Pekanbaru), Jambi, South Sumatra (Palembang), West Kalimantan (Pontianak) had hazardous air and PJJ — but no public BGN or media confirmation of MBG status in those specific cities during 14–23 Oct.

2. Whether Radar MBG internally logged haze status. The public view shows none. The backend may have a status field not exposed. Without access, I cannot confirm absence.

3. Whether SDMO issued internal circulars. The Oct 16 national mechanism (SIPERS-367) may have been accompanied by technical circulars to KPPGs that are not public. If they exist, they should be published.

4. Cooked-batch disposition data. The Oct 16 mechanism says already-cooked meals "distributed via accountable mechanism." No public record of how many batches, where, to whom, or with what documentation exists for any region.

5. Worker exposure metrics. No data on kitchen staff working hours, PPE provision, or health monitoring during the haze window.

Next move for MBG Watch

File a formal information request to BGN (via official channel and public letter) asking for:

1. The consolidated operating-status record for 14–23 Oct 2026 (or confirmation none was produced).
2. Radar MBG backend logs showing whether a haze/PJJ status flag exists and was used.
3. SDMO internal circulars issued 14–23 Oct on haze operations.
4. Per-SPPG cooked-batch disposition logs for Sumut, Aceh, Riau (Meranti), Jambi, South Sumatra, West Kalimantan, West Sulawesi (Mamuju).
5. Worker protection protocols activated (or not) during the window.

Timeline: Request sent → 14-day statutory response window → if unanswered or incomplete, escalate to Ombudsman and BPK as a transparency failure under the CMF's Elevated Concern tier.

Sources

1. Channel News Asia, "Singapore haze: PM2.5 hits 288, 12% schools closed," 14 Oct 2026 — supports Singapore threshold crossing.
2. BGN, Sidak SPPG Cadek, Kepala BGN Sudaryono Tekankan Kepatuhan SOP dan Akurasi Data (Pengumuman, 14 Oct 2026) — supports Oct 14 national publication (SOP inspection, no haze content).
3. BGN, Di Depan Kepala BGN, Pemda Aceh Buktikan Komitmen Dukung Program MBG (Siaran Pers, 14 Oct 2026) — supports Oct 14 national publication (regional monitoring urge, no haze content).
4. BGN, Mas Dar ungkap Mekanisme Penghentian Distribusi MBG saat PJJ Mengacu pada Pemberitahuan Pemda (Siaran Pers SIPERS-367/BGN/10/2026, 16 Oct 2026) — supports sole national haze-adjacent policy statement.
5. Antara News, "BGN: Mekanisme penghentian MBG saat PJJ mengacu pemberitahuan pemda," 16 Oct 2026 — corroborates BGN press release.
6. IDN Times Sumut, "Terdampak Kabut Asap Karhutla, MBG di Sumut-Aceh Dihentikan Sementara," 19 Oct 2026 — supports Sumut-Aceh regional halt 16–21 Oct.
7. Riau Pos, "Penyaluran Makan Bergizi Gratis Bagi 39 Ribu Warga Meranti Dihentikan Sementara," 20 Oct 2026 — supports Meranti halt since 9 Oct, 39k recipients.
8. Kompas Regional, "Sejumlah SPPG Mamuju Tetap Salurkan MBG Saat PJJ akibat Kabut Asap," 21 Oct 2026 — supports Mamuju 9/33 SPPGs halted 20 Oct.
9. Tribunnews Sulbar, "Alasan SPPG Mamuju Tetap Jalankan Sebagian MBG di Tengah Pembelajaran Jarak Jauh Karena Kabut Asap," 21 Oct 2026 — corroborates Mamuju detail.
10. MBG Watch, "Haze at the School Door: The Operating-Status Record MBG Needs Now," 23 Oct 2026 — the demand piece this briefing answers.
11. MBG Watch, "Continuous Monitoring Framework for MBG," 22 Jun 2026 — the tier definitions used for classification.
12. Rupiah Stability Watch, "The Convergence: Seven Operating Ledgers on One Meal Tray — November 2026 Update," 23 Oct 2026 — joint synthesis noting haze as one of seven converging ledgers.