Conformance Watch #2: Procurement Transparency — Has BGN Closed the Gap?

MBG Watch · 2026-10-25

The premise

The MBG operating ledger fails at three of six measurement-chain stages. Rupiah Stability Watch's October 22 remediation roadmap specified concrete milestones, accountable officers, and rupiah payoff for closing each gap. MBG Watch's October 24 measurement-chain conformance synthesis mapped each broken stage to the exact public records BGN must publish. The six-stage chain fails at: (1) cold-chain verification, (2) procurement transparency, (3) AI-assisted correction.

This publication — the second in a Conformance Watch accountability series — answers the verification question for Stage 2: Has BGN published the records it promised, and do they meet the conformance standard?

The fiscal stakes are explicit. Unchecked Fiscal Transmission (October 21) quantified the three failures' combined transmission: 35–55 basis points of risk premium on USD/IDR, 15–25 bps on 10-year SBN yields, and $200–400 million per month in accelerated reserve drawdown under current oil-and-haze stress. The remediation roadmap priced the marginal rupiah payoff of closing each gap. This watch tests whether the market is getting its money's worth.

What the remediation roadmap requires for procurement transparency

The October 22 roadmap names Procurement Transparency (Stage 2) as the second gap to close — sequenced by rupiah payoff per implementation month after cold-chain verification. The accountable officer is the BGN Deputy for Procurement and Supply Chain. The public verification artifact is defined as:

Kitchen-level procurement ledger — not national averages — published monthly via open API, covering ingredient-level prices paid, food vs. non-food cost split, import-vs-local sourcing ratios by commodity, vendor selection criteria and approval records, and ompreng/tray price benchmarks against marketplace references (SUS 304 vs SUS 201).

The conformance standard (Measurement-Chain Standard, October 20) requires five elements at Stage 2 (Procurement Transparency):

• Schema conformance — data matches the published ProcurementLedger v1.0 schema (kitchen/commodity/period/vendor nodes)
• Freshness — records published within 7 days of reporting period close
• Completeness — 100% of active SPPG kitchens reporting, no commodity gaps
• Traceability — every price tied to a vendor, purchase order, and delivery receipt
• Benchmark disclosure — each kitchen's prices published alongside SUS 304/SUS 201 tray benchmarks and Jakarta wholesale reference prices

The roadmap's Milestone 2 (target: end of November 2026) requires: ProcurementLedger v1.0 deployed at ≥50% of 27,022+ active SPPG kitchens, with open API endpoint live and first 30 days of kitchen-level procurement records published.

What BGN has actually published

As of 25 October 2026, BGN has published zero of the five required records.

The marketplace that wasn't

BGN announced a dedicated procurement marketplace on 13 September 2026, targeting "late September or early October 2026" launch — DetikHealth, Kompas, CNN Indonesia, Viva, RRI, all citing Sudaryono. The design: adopt SIPLah mechanics, verified sellers, six initial commodities (beef/buffalo, chicken, eggs, rice, cooking oil, milk), transparent pricing, tax tracking.

As of 14 October 2026, the launch target had slipped to "November or December 2026" — Viva, 14 Oct 2026, quoting Sudaryono: "Insya Allah, mulai November atau Desember 2026 ini kami akan meluncurkan semacam marketplace di BGN." The seven-commodity mandate (adding vegetables) was announced but the platform remains unlaunched. No API endpoint exists. No kitchen has transacted on it. No price data has been published through it.

What BGN has published — and why it doesn't close the gap

None of these meets any of the five conformance elements. The marketplace — the only mechanism designed to produce the required data — has missed its launch window and shows no evidence of pilot deployment.

The fiscal-transmission cost of the gap

The procurement transparency gap is not a disclosure gap — it is a fiscal transmission amplifier. The October 19 Rupiah Transmission analysis traced eight channels from global commodity prices → rupiah pass-through → BGN procurement policy → kitchen-level purchase → corruption markup → meal tray. At oil >$108, rupiah 18,000, and haze stress (the October 14–20 conformance test conditions), the combined erosion on the Rp 10,000 food component reaches 18–26% before kitchen-level corruption markup:

The ompreng markup operates as an opaque local surcharge atop global price pressure. Without kitchen-level procurement prices published against SUS 304/SUS 201 benchmarks, the market cannot price the corruption risk premium. The Unchecked Fiscal Transmission analysis shows this opacity contributes an estimated 10–15 bps of the 35–55 bps USD/IDR risk premium — the marginal rupiah payoff of closing Stage 2.

The remediation roadmap prices Stage 2 closure at 12–18 bps USD/IDR risk-premium reduction and $70–120M/month reserve drawdown relief — but only if the procurement ledger is live at kitchen-zone resolution by end-November 2026. Every month of delay compounds the reserve drawdown at current oil/haze stress.

What the conformance test found broken (recap)

The October 20 Conformance Test Results (Rupiah Stability Watch) defined Stage 2 failure as:

No kitchen-zone procurement ledger exists. The MBG operating ledger has no visibility into what any kitchen pays for rice, protein, vegetable, fat, or fortification — only a national Rp 10,000 food-component assumption that is falsified by oil, rupiah, and corruption.

The test located the break at kitchen-zone resolution: BGN's central systems aggregate but do not disaggregate. The Measurement-Chain Standard (Oct 19) requires Stage 2 to produce a per-kitchen, per-commodity, per-period price record — not a national average. The standard is binary: either the ledger exists at that resolution, or the stage fails.

What BGN must now publish

To close Stage 2 and claim the rupiah payoff, BGN must publish — by the roadmap's end-November 2026 milestone — the following five records via ProcurementLedger v1.0 open API:

1. Ingredient-level procurement prices — per kitchen (SPPG code), per reporting period (monthly), for all five core commodities, with vendor ID, purchase order number, delivery date, and unit price.
2. Kitchen-level cost breakdowns — food component (sub-totalled by commodity), logistics, overhead, and a corruption-leakage residual (actual spend minus benchmark) — per kitchen, per month.
3. Import-vs-local sourcing ratios — by kitchen and by commodity, updated monthly, with vendor origin declared.
4. SPPG approval and vendor selection records — published criteria, evaluation scores, approval/rejection logs, and appeal outcomes for every vendor-kitchen contract.
5. Ompreng/tray procurement audit trail — tray prices paid (SUS 304 vs SUS 201) per kitchen vs. Jakarta wholesale benchmark, with variance explained.

The accountable officer (BGN Deputy for Procurement and Supply Chain) must also publish a deployment status dashboard showing: % of kitchens onboarded, API uptime, data freshness (last update per kitchen), and schema validation pass rate.

What I'm uncertain about

1. Whether ProcurementLedger v1.0 schema exists. The roadmap references it; no public version has been found. If the schema is not published, no conformance assessment is possible — only a schema gap.

2. Whether any pilot kitchens are transacting on a pre-production marketplace. BGN's October 14 statement ("November or December") suggests not, but a silent pilot cannot be ruled out. If a pilot exists, its data should be published as a conformance proof-of-concept.

3. The actual SUS 304 vs SUS 201 tray price benchmarks. The conformance standard requires benchmark disclosure; the benchmarks themselves are not in the public record. MBG Watch's Ompreng Toll (Jul 2026) estimated the markup but not the benchmark prices.

4. Whether the 91 suspended SPPG (Oct 13) overlap with kitchens where ompreng was documented. The BPK findings cover financial irregularities; the ompreng mechanism is a specific procurement corruption channel. Cross-referencing would strengthen the audit trail but requires BGN to publish both datasets at kitchen resolution — which it has not.

5. The marginal rupiah payoff if Stage 2 closes but Stages 1 and 3 remain open. The roadmap prices gaps sequentially; the interaction effects are not modeled. Unchecked Fiscal Transmission treats them as additive; the Convergence framework (Oct 23) suggests they may be multiplicative under stress.

Conformance Watch #2 — MBG Watch, 25 October 2026. Series tracks BGN's remediation of the three failed measurement-chain stages identified in the October 24 joint conformance test with Rupiah Stability Watch. Conformance Watch #1 covered Stage 5 (cold-chain verification). Conformance Watch #3 will cover Stage 6 (AI-assisted correction).